]\gt ; IX .
Further Observations on itschia .
H. H. W. PEARSON , Sc. D. , P'.L.S .
Communicated by A. C. SEWARD , F.R. Received July Read [ PLATES 22\mdash ; 30 .
] TABLE 0F CONTENTS .
Page Introduction . . . . .
331 A. Inflorescence and Flowers . . . . . .
332 B. Microspore-mother-cell and Microspore 342 C. Pollinntion . . . . . .
343 D. Macrospore and . . . . .
344 E. end of Sac . . . . .
348 F. The lower end of the Sac the Endosperm . . . . .
351 Endosperms of Ephedra , Gnetum , and Welwitschia compared .
G. Germination of Pollen-grain and Growth of Pollen-tube . . . .
368 H. Fertilisation . . . . .
361 I. Development of Pro-embryo 364 K. ndosperm after Fertilisation 370 Summary of Results . . . . .
371 Disoussion . . . . .
371 Prothallus and Trophophyte . . . . .
375 ndosperm of Angiosperms . . . . . .
378 Pol rity of -sac . . . . .
381 Triple fusion . . . . .
383 ctoideae and Angiospermae . . . . .
385 -tnbes . . . . .
387 Distribution of etales . . . . .
388 Appondix I. . . . . . . . . . .
389 Appendix II . . . . .
390 Bibliogrnphy . . . . .
391 Explanation of Plates . . . . .
395 By the courtesy of His Excellency HERR VON LINDEQUIST and the Government of German South-West Africa , a second expedition to Damaraland was made the summer of 1906-7 .
Welwitschia was found in flower at Welwitsch and in the neighbourhood of Haikamchab .
The material which is the subject of this investigation was collected in these localities in January and February , The cost of the journey was defrayed by a grant from the British Association .
ARSON , ( b ) ; 1908 .
( See Bibliography at end of this paper .
) ( 270 .
) I am deeply indebted to Prof A. C. SEWARD , F.R.S. , for his advice on certain points discussed in the following pages , and for his kind assistance in co1mection with the publication of this paper during my absence from England ; to Miss STEPHENS , of Newnham College , Cambridge , for reading the proof ; and to other friends who have enabled me to overcome some of the serious difficulties occasioned by the inaccessibility of much of the recent literature dealing with this branch of investigation .
Material was fixed at various hours with a view to obtaining indications as to the effects of large differences of temperature and of atmospheric humidity upon the activities of the cell .
The best preparations , containing most of the important nuclear divisions , have been made from material fixed between 4.30 and 7.30 Speaking generally , the absence of contraction and the precision of the reactions to stains are less marked in oyules collected at other times of the day .
At the same time , nuclear division is far from being limited to the latter part of the day , and some good examples ( e.g. , figs. 77 , 78 ) were obtained between 1 and 2 , when the air temperature is near its maximum .
* In the early morning and late at night nuclear divisions are comparatively rare .
Only three preparations showing embryo-sacs with the free nuclei in simultaneous division have been obtained .
One of these is from a cone fixed in the field at Welwitsch between 1 and 2 .
It is to be Tetted that all information as to the time of gathering the ovules which yielded the two cases figured ( figs. 27 , 28 ) is unaccountably lacking .
These were fixed three hours after collection .
The preparation showing five of eight nuclei in mitosis in a later condition of the embryo-sac ( fig. 43 ) was obtained from an ovule collected at 6.30 , and fixed one hour later .
A. The Inflorescence Flowers ( Plate 22 , fig. 1 ) .
The unit of inflorescence , both male and female , is a dense spike or , as it is usually termed , conle .
The great profusion with which the cones are frequently produced is very remarkable .
In both cases they are normally borne in compound peduncled dichasia , usually more or less modified by arrest and sometimes by the development of more than branches at a node ; in some cases the arrest is so nearly complete that the inflorescence is represented by a single sessile or shortly peduncled cone .
The female inflorescence does not usually bear more than 23 cones which are sessile or nearly so ; in the male the branching is usually carried farther and the number of cones is proportionally greatel ; these are sessile or peduncled figs. Plate .
The occurrence of subfoliar inflorescences is far more common than was formerly supposed , and in plants large and small which in other respects appear to be *PEARSON , 1907 ) .
PEARSON , 1907 .
HooKEIt , 1863 , p. 20 ; PEARSON , 1906 , p. 272 .
PROF .
H. H. W. PEARSON : FURTHER OBSERVATIONS ON WELWITSCHIA .
333 perfectly normal .
In some cases successive fertile ridges are formed under the leaf in acropetal order and with almost as much regularity as the axillary ridges in normal plants .
When this happens the subfoliar inflorescences are commonly more abundant than those in the ordinary position and , if not confined to one side of the plant , they are never equally developed on both .
As a result of this abnormal lateral growth the crown with its leaves is displaced and may turn vertically through an angle as great as These inflorescences do not arise irregularly from the older part of the stock as in some woody angiosperlns , but the abnormality is due rather to the appearance of a floriferous meristematic zone below the leaf as well as , or instead of , in its axil .
here is no dicatio that this is in any way directly influenced by external conditions .
One isolated valley near Haikamchab contained 106 plants , of which 24 had not yet flowered .
The youllgest of these was certainly older than the , figured in i HOOKER'S Monograph ( Plate 2 , fig. 1 ) .
This was the smallest found , and , judging by such standards as are available , could not be less than 10 years old ; it bore a single arrested cone of doubtful sex .
Seyeral plants llot much more advanced were in flower , and it seems probable from a study of these that the male flowers earlier than the female .
This was the only locality visited in which the male plants were to any marked extent more numerous than the female .
The female cone has , at the base , a pair of bracts distinguished from those above them by their acute apices fig. .
Frequently a considerable length of axis separates them from the rest of the cone .
Exceptionally the one or two pairs are of like character .
These are either barren or bear in their axils arrested or normal cones , or cone-bearing branches .
The bracts above these become successively broader and have a rounded or ver slightly acute apex fig. .
The lower of these\mdash ; usually about 20 , frequently as few as 12\mdash ; are barren and smaller than those above them .
All the upper bracts are fertile .
Sixty flowers have ) eell found in a single cone and this number is probably but rarely exceeded .
All ) of development may be present in the same cone .
Oyules situated around the apex become arrested before they the stage at which pollination occurs .
These with their subtending leayes form a withered tuft .
the otherwise axis after the fall of the seeding flowers and their bracts -r As in the female , a pair of acute bracts stand at the base of the male cone and subtend perfect or arrested cones or cone-bearing branches .
The next pair or two * pairs of bracts are barren ; all those aboye have flowers in xils .
number of flowers nuay be as great as 70 , and is usually about 50 .
A at the top of the cone do not reach maturity .
Macrospores and embryo-sacs up to a stage containing about 64 , nuclei are RSON , , fig. , 1863 , PIate 8 , fig. 1 .
: * ; PROF .
H. H. W. PEARSON : FURTHER OBSERVATIONS ON WELWITSCHIA .
with one integument inserted on an ovuliferous scale of which no traces remain .
This view appears to be based entirely upon analogy with the Coniferae .
It is not supported by the anatomical structure and development of the female spikes of these genera as interpreted by other investigators .
There are indications that the compound inflorescence in Gnetum is derived from a simple spike .
As in Welwitschia , there is at the the female spike a pair of opposite bracts .
* These usually subtend a pair of axillary spikes ; the buds which normally give rise to the latter may , however , relnain dormant ; sometimes they develop into normal flowers .
It appears , therefore , that in Gnetum the lateral spike is equivalent in origin and position to a flower .
In Welwitschia the evidence is less complete .
Attention has already been directed to the two acute bracts at the base of the cone and to the varying degrees of development of their axillary buds which , however , so far as is known , never produce normal flowers .
It is suggested that the dichasium is derived from the cone by tion , and that the lateral cones directly replace flowers , as is known to be the case in Gnetum .
According to this hypothesis the axes of the lateral cones are nucelli of theoretically indefinite growth , and the presence and acropetal arrangement of embryo-sacs within them is to this extent explicable .
It is unfortunately impossible at present to determine whether the terminal cone of the Welwitschia dichasium ever contains these extra-floral macrospores .
They have been found to be absent in some cases , but in the material available the terminal cones cannot be distinguished .
But they occur in so large a proportion of those examined , and the terminal cones are so compared with the lateral , that even if they are present in the former they cannot confined to them .
It is to be expected that they will be found also in the lateral cones of Gnetum .
Their absence from the axis of the male cone , if rmed , is not remarkable on this hypothesis , for the oyule of the male flower apparently never produces sporogenous cells .
The morphology of the male and female flowers of Welwitschia has been discussed and very diverse conclusions have been arrived at by different There are , in the male flower , a well-marked androecium and an arrested ovulal body which all admit to be the homologue of the functional nucellus of the female .
But the ovular envelope and the two lower whorls of appendagres of the male and the two envelopes of the female have been riously interpreted .
views authors are given in Table I , in which the numbers indicate the of appendagesS from below upwards .
These divergent views are the outcome of attempts to the one flower from the other or both from a hermaphrodite , assumed to be more less represented * ENGLER , 1908 , p. 520 , figs. , K. EICHLER , 1887 , fig. 76 .
LOTSY , 1899 , p. 85 , PIate 2 , fig. 6 .
S See pIlARSON , 1906 , text-figs .
1 and 2 .
PROF .
H. H. W. PEARSON : FURTHER OBSERyATIONS ON WELWITSCHIA .
337 respectively on the receptacle of many Compositae .
While most authors agree with WIELAND* in regarding the seed-pedicels as foliar structures , the point is not definitely settled , and it may therefore be permissible to compare the ovulate strobilus of Bennettites with , the female cone of Welwitschia .
The Welwitschia cone ( 2 ) shows an elongated axis whose vegetative apex remains active to the last .
Upon it are borne from below upwards ( see Table II ) : ( 1 ) one ( sometimes two or three ) pairs of acute bracts ; ( 3 ) about 10 pairs of imbricating bracts increasing in size from below upwards fig. ) ; ( 4 ) bracts , similar to ( 3 ) but larger , bearing in their axils sessile or very shortly stalked female ' ' flowers This differs from the ovulate strobilus of Bennettites in that ( 1 ) the hypogynous disc of microsporophylls is absent ; ( 2 ) the bracts are in whorls of two instead of being arranged spirally ; ( 3 ) the upper part of the cone bears only one bract to each flower , instead of five six .
The comparison may be extended to the ovules .
In both cases the ovule is orthotropous and surrounded by a single integument which is prolonged upwards into a micropylar tube , the only ovular structure to project beyond the bract .
Outside the integument in Bennettites is " " cup-shaped supporting basal husk " " formed of " " cortical cells several cdls deep at the base and thinning out towards the A characteristic feature of the -called " " peria ) \ldquo ; of Welwitschia is the great development in the lateral wing-like extensions of its wall of tubular cells , long , tortuous , and " " stringy.\ldquo ; It is , however , a much more highly organised structure than the cupule of Bennettites , for , in addition to its form , which indicates all adaptation to the purpose of seeddistribution , it contains a vascular tissue .
There are also internal characters which ossibly indicate an even more striking resemblance between these ovules , and which will be referred to in a later page ( see p. 386 ) .
As to the male flower , WIELAND remarks , S " " nor is there from a plain point of view an unbridgeable gap between the staminate disc of Cycadeoidea and that of Welwitschia , for the latter could arise similarly to one of the hypothetical one-seeded and bisporangiate forms of the Cycadeoidean alliance by one of the simplest of all evolutionary processes , namely , increased number of flowers to the plant and decrease in bulk until there was left of each original frond but a single filament bearing . . .
pollen sacs The cone itself might be derived from the flower by a process of proliferation It appears , therefore , that if the ovule of Bennettites is axial it is theoretically possible to derive both the female cone and the male flower of Welwitschia from an strobilus of this type .
The former requires the abortion of the *Cf .
ARBER and PARKIN , 1907 , pp. 55 and WIELAND , pp. 120 , 121 .
HOOKER , 1863 , Plates 7 and 8 ; BURGE1 , 1872 , rfaf .
18 , fig. 22 .
S WIELAND , 1906 , p. 245 .
WIELAND , 1908 .
VOL. CC.\mdash ; B. 2 disc of microsporophylls\mdash ; a tendency towards which is perhaps shown by the fossil forms themselves\mdash ; and minor changes in the relative growth of the main axis and of the lateral members , and in the number of the sterile appendages as compared with the fertile lateral axes .
The latter demands reduction and specialisation in the microsporophyll-whorl , the disappearance of the interseminal scales , the reduction of the ovulate axes to one terminal arrested ovule and the abortion of one of its envelopes .
The suggested relationships of the various parts are expressed in the following tables .
This view seems to accoun .
more naturally than any hypothesis yet advanced , not only for the separated sexes , but also for the relations between the parts of the male and female flowers .
The ovules in both flowers are clearly caulin and it must be assumed that they have come from a type in which this was also a character .
Ancient types , in which the ovule was apparently axial , are known , and , if the comparison with Bennettites is invalid , it may still be that the relations between the Welwitschia cones are of the nature suggested , , that the female cone and the male flower are derived by two independent series of reductions from a primitive amphisporangiate strobilus with axile ovules .
Pendin a detailed anatomical investigation , the internal structure of the Table II .
rren inter.arren baxisp yorsporop ycaducous , ftcormsoweonhich s *WIELAND , 1906 , text-figs .
56 , 87 , 88 .
PEARSON , 1906 , text-fig .
1 .
See Table III .
PROF .
H. H. W. PEARSON : FURTHER OBSERVATIONS ON WELWITSCHIA .
339 Table III .
envelopes .
BenIlettitean ovule and flower .
\mdash ; WlELAND , 1906 , text-fig .
63 .
PEARSON , 1906 , tcxt-fig .
2 .
PEARSON , 1906 , text-fi cone-axes , .
bracts , and floral appendagesS may be briefly compared .
The vascular structure of the peduncle of the female cone has been described by STRASBURGER and by The axis is occupied by a parenchymatous pith containing a few sclerotic cells and elongated " " spicular\ldquo ; elements .
Tn the peduncle of the male cone the bundles of the inner ring border upon an equally massive pith of a different character .
The cells , of which it is almost entirely composed , are from six to twelve times as long as they are broad ; they are uniformly arranged , with their axes parallel with that of the peduncle .
Their walls are nified and contain very numerous circular or oval simple pits .
They are broader and shorter than the tracheids of the xylem , but perhaps represent a more extensive development of the sclerotic cells in the pith of the female peduncle , which WoRSDHjLL , " " although with some hesitation inclined to regard as the representative of centripetal xylem .
Mixed with them are a few spicular cells , which are less common than in the corresponding region in the female .
The vascular system is more abundantly developed in the female than in the male .
The acute bracts which stand at the base of both cones are also alike in structure , though differing in form .
The bundle is endarch and the xylem appears in transverse section as a narrow plate lying along the inner face of the phloem .
Radiating from it S I am indebted to Mr. H. BOSMAN and Mr. P. .
MOLL the prep ration of series of sections of these structures .
STRASBURGER , 1872 , Taf .
20 , figs. 49-51 .
WORSDELL , 1901 .
PROF .
H. H. W. PEARSON : FURTHER ON WELWITSCHIA .
341 avoid the conclusion that in these forms the ovule is not merely an placed with a distinctive tegumentary system , modified bud .
This view , if correct , would seem to point to the conclusion that the ovule throughout the group , whatever its position , is line .
On the other hand it is , perhaps , not Ilecessarily regarded as in all cases to the same category .
But the point of immediate interest is that in some ngiosperms reasonably regarded as relatively primitive the ovule is apparently an axial structure , as in Welwitschia .
There is therefore suggested in Tables II and III a comI ) arison between the floral members of such an angiosperlu and those of Welwitschia .
If the relations which there appear can be maintained , neither of the Welwitschia flowers possesses a structure homologous with .
A possible alternative is that the outer winged envelope of the female flower is a primitive ovary\mdash ; as has already been ested by PARLATORE and authors ( see Table I ) .
In this case the ovular axis produces the ovary as a pair oflateral appendages\mdash ; a view which is very difficult to reconcile with the relations between the ovule and the carpel , wonly obtain among the angiosperms .
It seems , therefore , more probable that if the carpel is represented at all in the ovulate cone of Welwitschia it is by the bract which subtends the ovule .
On this hypothesis , if we imagine the cone so much reduced that the arrested axis bears only two bracts and a single ovule in the axil of one of these , we arrive at a structure closely resembling the young flower of reduced inflorescenctl , in the same sense that the female cone of Welwitschia and , as suggested above , the ovulate portion of the Bennettiteall strobilus , are inflorescences .
At a later stage the elnbryo-sac of Peperomia will be coml ) ared with that of Welwitschia , and the esults will tend to confirm conclusions tlJat it represents a ) rimitive type of angiosperm sac .
Many writers have maintained that Peperomia is a educed form both as its vegetative its floral structures .
S these views are not incompatible with of the embryo-sac , the new light which Welwitschia throws upon the latter is rather favourable to relatively low position usually to the genus by recent xonomists .
ARBLR have advanced The most important of these are : ( 1 ) that , on the of the origin of flower a Bennettiteoid strobilus , it presupposes that the perianth arises de noro , and ( 2 ) that eromia { forms with simple apetalous flowers possess a and evolved inflorescence .
first objection raised is undoubtedly serious if ngiosl ) erlns , and flower ) *BALFOUR , 1901 , p. Cf .
ICHLER , WORSDELL , 1904 .
, 1900 , fig. 3 .
S .
T. G. , 1906 ; ARBER PARKIN , 1907 .
ENGLER , 1907 ; also A. ) , 1906 , 1 ) .
425 .
ARBER and PARKIN , 1907 .
342 PROF .
H. rr W. PEARSON : FURTHER OBSERVATIONS ON WELWITSCHIA is derived from what the authors term the " " Proanthostrobilus But while so little is known of the life-histories of so many angiospermic families , and the phylogeny of the group is so obscure , evidence tending to show that Peperomia and other apetalous forms are relatively primitive is not invalidated by the presence of a perianth in other groups .
On the " " proanthostrobilus\ldquo ; hypothesis , it is certainly not easy to account for the dense spike of a primitive Peperomia .
But too much importance may easily be attributed to this point .
The male cone of Welwitschia\mdash ; really a dense spike of reduced hermaphrodite flowers\mdash ; presents no less difficulty , and is very considerably more primitive than either the spike of Peperomia or the flower of a Magnolia .
Physiologically the integument of Welwitschia is to some extent comparable with the ovary of the angiosperm .
Not only does it receive the pollen , but it also furnishes some , at least , of the nutrient medium in which its germination and , frequently , the early growth of the pollen-tube takes place ( fig. 11 ) .
The outer envelope of the fertile ovule performs two of the functions commonly associated with the ovary of the higher plants , , the protection of the ovule and the distribution of the seed\mdash ; for which its winged form is admirably adapted .
* In this it is assisted by the broad subtending bract to which the seed remains adherent for a time after the disintegration of the cone .
B. The Microspore-mother-cell and the Microspore ( Plate 22 , figs. 2-7 ) .
The microsporangium formerly contains pollen-mother-cells in the resting condition .
The loculus undergoes a further increase in size and the cells of the inner layer of the tapetum become considerably larger .
At this time they are almost constantly bi-nucleate and their cytoplasm is very scanty .
The outer tapetal cells ( m.w. ) are very much flatten.ed and the three wall-layers now resemble those Ephedra in a somewhat later stage .
S The breaking down of a number of the spore-mother-cells seems to be a normal occurrence : not infrequently the loculus contains nothing but the disorganised remains of the whole cell-mass .
The wall of the mother-cell shows no appreciable increase in thickness during its development .
The number of the chromosomes in the heterotypic division lies between 22 and 26 , and is believed to be 25 .
The spindle is at first multipolar , and in one stage closely resembles a condition figured by MOTTIER for Tradescantia The stout fibres of the bipolar spindle show slight curvatures in opposite directions at the poles ( fig. 3 ) When the daughter-nuclei are organised there is seen to be no cell-plate between them ( fig. 4 ) .
, 1863 , Plate 7 , fig. 3 .
, fig. 2 .
PEARSON , 1906 , fig. 4 .
S .
LAND , 1904 , fig. 14 .
, 1907 , fig. 40 .
PROF .
H. H. W. PEARSON : FURTHER OBSERyATIONS ON WELWITSCHIA .
343 During the homotypic divisions the cytoplasm possesses a greater density than in the preceding mitosis , and this increases up to the organisation of the microspores .
The spindles are frequently placed at right angles to one another ( fig. 5 ) , but they may be inclined at any angle , and sometimes their axes are parallel .
Usually the daughter-nuclei show the tetrad arrangement ( fig. 6 ) ; sometimes a section passes through them all .
In addition to those of the two spindles , fibres appear between the two pairs of daughter-nuclei and separating walls are formed in the usual manner ( fig. 7 ) .
At the same time a new wall is formed in the periphery of the cytoplasm , and there result four free microspores still enclosed in the wall of the mother-ceI The account already given of the structure of the adult pollen-grain is confirmed , But the nuclei are not confined to the positions nerly indicated ; they have frequently been seen together at one end of the cell , and sometimes the generative cell and tube-nucleus are separated by the whole length of the cell .
It is therefore possible that in the living pollen-grain they may be situated in any part of the cytoplasm ; but the thick spore-wall no doubt prevents rapid fixation , and the cytoplasm is always more or less contracted .
C. Plate figs. .
Pollination is mainly effected by the hemipterous insect The pollen is received by a drop of a sweet fluid which stands on the top of the projecting micropyle when the oyule is ready for fertilisation .
The secretion of this fluid undoubtedly takes place largely in the cells the inner lining of the integument and micropyle figs. ) and the outel covel.ing of the nucellar cone figs. page .
Fig. esents a section of these and neighbouring layers in a young ovule not yet ready for pollination .
In fig. 10 is shown a transverse section of the same layers in an ovule which contains young embryos\mdash ; i.e. when the secretion has ceased .
The greatly enlarged cells of the layer , , would lead to the conclusion that it plays a part in the function .
Both layers , together with the outer cells of the umento , of an ovule at the of pollination give a copious reaction for cane-sugar with Fehling 's solution .
S The appearance of sugar is accompanied by a depletion of the starch which , in the stage of fig. 9 , is contained in several hypodermal layers of the nucellar cells ( cf. 10 ) .
The excretion of the sugar solution only commences after sunrise and ceases very soon after , if not before , sunset It was not mined how long all * PEARSON , 1906 , fig. 11 . .
cit. , fig. 16 .
This is in direct contradiction of a statement formerly made in error : see PEARSON , 1907 .
S Only dried material is available for this reaction ; it is quite ) that other regions of the nucellus , internal to those indicated , are also COIlcerned in the sccretion of the micropylar fluid .
PEARSON , 1907 .
344 PROF .
H. H. W. PEARSON : FURTHER OBSERyATIONS ON WELWITSCHIA .
ovule remains in the condition for pollination , but a drop probably appears on the top of the micropyle day after day for two or three days at least .
The ovule frequently receives so large a number of pollen-grains that thsy fill a considerable length of the micropyle ( fig. 11 ) .
It is probable from their position and number that they sink in the fluid by reason of their own weight .
Germination occurs while the pollen-grain lies on the top of the nucellar cone*or , very frequently , at some distance above this in the micropylar fluid ( fig. 11 ) .
This is to some extent comparable with the growth of the pollen-tube down the style of an angiosperm , but in Welwitschia the nutrient medium in which it grows is partly derived from the nucellus .
D. The Macrospore the ( Plates 23 and 24 , figs. 12-23 ) .
An ovule appears never to contain more than a single macrospore-mother-cell .
In spite of prolonged search only one stage of the heterotypic division ( fig. 12 ) has been found .
It is considerably later than that formerly figured and is apparently a stage in which the spireme is recovering from synapsis .
This condition has been seen several times and it is probable that other stages are passed through with greater rapidity .
The spireme presents a beaded appearance owing to the presence in it of very numerous granules , irregular in form and size , staining faintly in diamant fuchsin .
Here and there a longitudinal splitting of the thread is suggested .
The division of the daughter-cells has not been seen .
Since three embryo-sacs are occasionally present it may be inferred that at least the lower one undergoes a homotypic division .
In general only one macrospore germinates .
Two have been found in a few oyules , most of which were taken from the same cone .
Tn these cases they usually lie one above the other , the lower one , which survives , being more advanced in development ; rarely are they situated side by side .
In one ovule three macrospores were germinating : two were side by side and equal in ] opment and the third above them less advanced .
The macrospore is usually broader at the base than at the apex ( fig. 13 ) , and these relations are commonly maintained up to the time of septation .
The wall is thick ; it shows no signs of cutinisation .
The cytoplasm , in which the large nucleus occupies an approximately central position , contains numerous small vacuoles fairly uniformly distributed through it .
The late prophase of the mitosis of the macrospore-nucleus is shown in two sections , oblique to the axis of the spindle , in fig. 14 , A and B. The arms of the spindle are slightly curved in opposite directions at the poles ; the fibres are faintly stained in * PEARSON , 1906 , fig. 17 .
PEARSON , 1906 , fig. Cf .
MOTTIER , 1907 , PROF .
H. H. W. PEARSON : FURTHER OBSERVATIONS ON WELWITSCHIA .
erythrosin .
At each pole is a vacuole distinguished from the rest by its definite outline .
The chromosomes are small ; their is not less than 22 nor more than 26 .
The daughter-nuclei are formed not far from the poles of the cell ; there are no constant visible ' differences between them in form , size , or staining capacity .
They have not been found in division , but in the next stage ( fig. 16 ) the nuclei are always approximately equidistant and the slight polarity of the sac has disappeared .
is a point of some interest , for it has been suggested that in Gnetum the descendants of one of the two nuclei of the stage of fig. 15 the micropylar ) are all sexual in function , while those of the other are sterile*\mdash ; view which naturally leads to a comparison with the micropylar and antipodal groups respectively of the angiosperm-sac .
The distribution of the nuclei in this early stage ( fig. 16 ) of the Welwitschia sac offers no support to this hypothesis ; on the all their visible characters indicate that there are no fundamental differences , actual or potential , between them .
Further evidence confirmation of this conclusion will become available later .
In the bi-nucleate sac figured ( fig. 15 ) the cytoplasm contains , near the periphery and almost midway between the nuclei , a spherical vacuole with a very definite boundary lined with small irregular granules staining brilliantly in erythrosin .
This bears some resemblance to the polar vacuoles of fig. 14 and is probably a structure of the same kind .
It is not a constant feature of this condition of the sac , nor has it been observed in a later stage .
Each of the four nuclei in fig. 16 is surrounded by dense cytoplasm , outside which there is a marked development of small vacuoles .
The absence of a large central vacuole , which was formerly stated to be a constant feature of the free-nucleate condition , is confirlned .
This can only mean that , in proportion to the increase in the area of the macrospore-wall , rowth of the cytoplasm proceeds more rapidly than is usual among Gymnosperms .
Whetller this is due to an abnormally rapid formation of new cytoplasm , or to an unusually slow increase in the peripheral area of the sac , or to both , is not certain .
The small size of the sac at the time of its septation , together with the fact that the vacuole is similarly absent in Gnetum when the normal growth of the sac is hindered , favours the second view .
Other evidence , however , makes it probable that the absence of the vacuole is , in part at least , directly due to the rapid growth of the cytoplasm and ication of the nuclei .
In a cone , most of whose ovules ah.eady contain embryos , a few near the apex are still in early of development .
In these the free nuclei always show tendency to confine themselves to a lining layer of and deyelopment of vacuoles is much greater than in corresponding stages in younger cones ( fig. 17 ) .
This appeal .
S to be a natural consequence of the etative activity of the old cone in *LOTSY , 1899 , p. 92 .
PEARSON , 1906 .
ICARSTBN , 1892 p. 212 .
VOL. CC.\mdash ; B. 2 PROF .
H. H. W. PEARSON : FURTHER OBSERVATIONS ON WELWITSCHIA .
347 and 1098 .
Assuming the constancy of simultaneous divisions up to this point , the actual number is 1024 .
In other words , the nuclei present in the sac when septation begins all belong to the 11th generation from the nucleus of the macrospore .
Between these two stages ' ( figs. 20 , 21 ) the vacuoles have become greatly reduced , and the nuclei in the later one are very crowded .
Each shows , as in earlier stages , a single nucleolus .
Very soon after the attainment of this condition the whole sac undergoes a rapid increase in length .
The cytoplasm again becomes vacuolated and the nuclei less crowded ( figs. 22 , 23 ) , while the sac itself attains a length three times that of fig. 21 outlines .
In the upper fourth fig. the nuclei are less crowded , and the cytoplasm more vacuolated than in the lower three-fourths ( fig. 23 ) ; it may be inferred that the upper end of the sac increases in length more rapidly than the rest .
This is also indicated by the fact that , while the antipodal end still lies free in the products of decomposition of the nucellar cells , the micropylar apex has pushed its way into the midst of nucellar tissue whose continuity remains unbroken .
These changes and those which immediately follow them are effected with great rapidity and stages intermediate between those of figs. 21 and 22 , 23 have rarely been seen ; while they are in progress the whole sac is exceedingly liable to contraction during preparation : As in the sac of fig. 21 , so in that of figs. 22 , 23 , the nuclei themselves are alike in all visible characters ; in intermediate stages , those of the apical fourth show a solnewhat greater avidity for stains than the rest , but this difference persists only for a short time .
The nuclei of figs. 22 , 23 are crowded with chromatin granules to such an extent that the nucleoli are obliterated ; in some other preparations of similar age a single nucleolus in each is recognised with difficulty .
sac now shows for the first time a clear differentiation into two regions , the one fertile and the other entirely , or for the most part , sterile .
The more widely separated nuclei of the upper fourth ( fig. 22 ) are functionally sexual ; of the more crowded ones of the lower three-fourths ( fig. 23 ) , all , or the great majority , will presently give rise to a sterile nutritive tissue , the endosperm ; a few of the latter however frequently remain functionally sexual and behave exactly like .
those of the fertile end .
The fertile nuclei of upper end are few compared with those which form the sterile tissue .
It follows , therefore , that if the simultaneous divisions are maintained up to this stage\mdash ; and there is no evidence that they not\mdash ; the suggestion referred to above , that the sexual nuclei are all descended from the one and the remainder all the other of the two daughter-nuclei of macrospore-nucleus , is completely disproved .
The cytoplasm of the section from which figs. 22 and 23 are taken is already marked out into areas by tortuous lines , which stain deeply in lich -griin ; these are sections of plates of denser cytoplasm , the boundaries of compartments , into which the sac is subdivided .
These initial planes of septation bear no visible relation to the spindles formed in the last division of the free nuclei ( cf. figs. 21 , 22 , 23 ) ; and the sac undergoes so considerable an amount of growth , resulting in the wider 348 PROF .
H. H. W. PEARSON : FURTHER OBSERyATIONS ON WELWITSCHIA .
distribution of the nuclei between these stages , they are almost certainly independent of nuclear division .
But ce an enucleate compartment has never been seen , it is probable that their positions are in definite relation to the distribution of the nuclei .
Their formation is probably the result of a process of cleavage similar to ' that described for certain Myxomycetes and Phycomycetes , * and resembling in some respects that which precedes the organisation of the pro-embryonal cells in the egg of The stage immediately preceding that shown in figs. 22 , 23 has not been seen , and it has not been determined whether the formation of these cleavage planes is centripetal or whether they appear simultaneously roughout the sac .
Near the periphery the cleavage lines usually end in depressions or deep furrows in the cytoplasmic surface .
It is therefore possible that they commence , as in the of Synchitrium , in these superficial furrows and advance centripetally .
However this may be , the of the micropylar cytoplasm occurs in exactly the same way as that of the lower part .
It appears from fig. 32 , in which the cytoplasm is probably contracted , that the cleavage planes of fig. 23 really consist of two asmic surfaces in contact ; here are slightly separated by a clear space , in which no trace of a cell-wall can yet be detected .
Between these surfaces the cell-wall is formed and , in the stage of figs. 25-33 , reacts to cellulose stains .
The compartments of the micropylar fourth are somewhat larger , and include fewer nuclei and larger vacuoles than those of the lower three-fourths .
In the former the nuclei are frequently not more than two , only exceptionally more than figs. ; in the latter their number is variable and , owing to the irregular form of the compartments , not easily determined ; it is usually twelve or more , occasionally as small as two .
There is no indication that any factor is concerned in bringing about this unequal distribution of the luclei except the unequal growth in length of the sac after the stage of fig. 21 ( see below , p. 354 ) .
The nuclei themselves , however , are of the same generation , and are equal in every respect , save only that there may still be retained some traces of the difference in staining capacity referred to above .
All the nuclei of the stage of figs. 22 and 23 are potentially sexual .
From this onwards the two ends of the sac develop along different lines , and .
later histories are most conveniently considered separately .
E. End of the Sac ( Plates 24 and 25 , figs. 24-31 ) .
The nuclei of the micropylar compartments increase in size and the single nucleolus , once more distinct , is situated in a large clear space ( figs. 24 and 25 ) .
Those of the of the sac later undergo the same changes ( cf. figs. 32 , 1899 , hIoTTIIlR , 1904 , p. 36 .
LAND , 1907 , figs. 6-9 .
The statement formerly made PEARSON , , that ' unlil considerably more advanced stage more than two nuclei are llot found\ldquo ; in these compartments , is incorrect , sometimes happens that a downward direction is followed ( fig. 27 ) , and even more frequently they turn sharply outwards and pass more or less horizontally towards the integum nt ( fig. 10 ) , and may even enter the space between the latter and the side of the nucellar cone .
Each tube grows independently of the rest .
Of the majority which take an upward direction each continues its growth until a pollen-tube is encountered and the fertilisation of one of the female nuclei is effected .
Unfertilised tubes continue to grow until they reach the surface of the nucellar cone .
As they advance beyond the normal region of fertilisation they broaden and their cytoplasm becomes much vacuolated .
As KARSTEN* has stated , they may even emerge from the nucellar tissue into the micropyle .
In the early condition of the tubes the nuclei lie close together ( figs. 26 , 28 , 29 ) .
Later they usually become separated and are arranged in single file ( fig. 31 ) .
When the embryo-sac-tube meets the pollen-tube the female nucleus is commonly considerably in advance of those behind it ( e.g. , fig. 53 ) .
During the close aggregation of the nuclei in the earlier stages irregularities of form are usually impressed upon them and these are frequently retained up to the time of fertilisation ( cf. fig. 31 ) .
In view of the fusion of the nuclei which normally occurs in the multinucleate compartments of the lower part of the sac , the fact that an entirely similar fusion of all nuclei of an embryo-sac-tube may exceptionally occur is of very great interest .
Two such cases are shown in figs. 30 , A and B. In the former about six nuclei seem to have participated in the fusion ; in the latter the number was probably smaller .
The nuclei of fig. 31 are ready for fertilisation .
It is to be emphasised that no nuclear divisions have intervelled between the stages of figs. 22 and 31 ; in other words , the functional female nuclei , although differing in size and to some extent also in staining capacity , are identical with those which occupy the micropylar end of the sac of fig. 21 .
In most cases there is nothing to indicate that the cytoplasm immediately associated with the nuclei of the embryo-sac-tube has undergone any special degree of organisation ; rarely , however , a thin envelope stains more deeply than the general cytoplasm of the tube ( fig. 55 , Plate 27 ) .
There is , however , no sharp lin separating the one from the other , and it seems that , as in Gnetum , the female gametes are free nuclei .
Fertilisation may take place in any part of the nucellar cone and in rare cases it may perhaps occur in the axis below the insertion of the integument .
But the normal region of fertilisation is situated not far below the line of the text-figure .
Since the embryo-sac-tubes have undergone a considerable growth before pollination ( e.g. , fig. 27 ) , the sexual fusion cannot normally occur within the limits of the embryo-sac itself , and no indication that this ever happens has been met with .
While in its general behaviour the embryo-sac-tube closely resembles the pollen* KARSTEN , 1893 .
See also PEARSON , 1906 , figs. , 34 , 40 .
PROF .
H. H. W. PEARSON : FURTHER OBSERVATIONS ON WELWITSCHIA .
351 tube , in one respect there appears to be a remarkable physiological difference between them .
The pollen-tube contains a vegetative nucleus\mdash ; the " " tube-nucleus\ldquo ; \mdash ; usually situated not far from its tip .
This normally shows signs of commencing disintegration shortly before the elongation , of the tube ceases .
While the growth of the tube may continue for some time after its disappearance ( figs. 47 and 48 ) , it is nevertheless probable that it performs a definite directive or nutritive function .
In the embryosac-tube , on the other hand , all the nuclei are apparently equivalent and that which lies nearest the tip normally becomes fertilised ( figs. 31 , 53 , 55 , 58 ) , and there are no signs that it is more liable to become degenerate than are those which follow it .
It therefore appears that the growth of the tube is independent of a vegetative nucleus .
F. The Lower End of the Sac tloe Endosperm ( Plates 25 and 26 , figs. 32-39 ) .
The marking out of the lower three-fourths of the sac into compartments\mdash ; for the most part multinucleate\mdash ; and the laying down of cell-walls in the planes of septation take place as in the upper end .
It is possible that these processes occur a little earlier in the micropylar end ( cf. figs. 24 and 32 ) .
The lower compartments are less vacuolated and , on the whole , smaller than those which contain the fertile nuclei .
The irregularity of form characteristic of them when first established becomes less marked in later stages ( cf. figs. 23 and 33 ) .
The nuclei , at first with the nucleoli more or less obscured by chromatin granules ( fig. 23 ) , slowly increase in size and the nucleolus becomes once more visible .
As the staining capacity of the wall increases ( fig. 33 ) , the nuclei attain their full size and become identical in appearance and reactions with the functionally sexual nuclei of the same sac ( fig. 25 ) .
One free nucleus of fig. 33 and all those of one compartment of fig. 34 are in mitosis ; in the latter case the eight nuclei , five of which are shown , are in the same stage of division .
These are the only instances observed of the division of the free nuclei of the sac after the inception of septation , and it is clear from a comparison of the numbers in the sacs of figs. 21 , S3 , 34 , that it is not of general occurrence .
A similar simultaneous division of free nuclei occurs in the septate endosperm of there , howeyer , it appears not to be limited to a few compartments but to extend rough all .
Whatever may be the significance of these divisions , they are comparatively rare in Welwitschia .
The nuclei in each compartment now approach one another and accumulate near the centre ; in this condition , except for their number , they closely resemble the sexual nuclei of fig. 25 .
Here , however , they do not remain distinct : at the places of contact the membrane disappears and a gradual fusion occurs ( fig. 35 ) ; each compartment in this way becomes a uninucleate cell ( tigs .
35 and 36 ) .
The fusionnucleus usually retains several nucleoli until its first division occurs ; in other STRASBURGXR , 1880 , Taf .
2 and 3 , figs. 56-60 .
: cases a large nucleolus , apparently resulting the fusion of several smaller ones , is present .
free nuclei in the incompletely septate sac of Corydalis fuse in a precisely nilar manner to form the nuclei of the endosperm cells .
The fusion of the nuclei of the multinucleate compartments of the sterile region of the sac must be effected with considerable rapidity .
As a rule , in a cone bearing 40 or more ovules , the earlier stages shown in figs. 35 , , are to be found only in two or three .
) absence of similar stages from the material collected in 1904 is no doubt to be accounted by the period which elapsed between its collection and fixation .
A few cases of the last stages of fusion were figured ; these all occurred near the normally fertile region of the sac and , being associated with the close approximation of the female nuclei in the fertile compartments and the early condition of the embryo-sac-tubes , were 1nisinterpreted as stages in amitotic division from which , it was erroneously suggested , the sexual nuclei resulted .
S Frequently the fusion-nucleus undergoes an immediate mitosis figs. .
Whether it occurs at once or is delayed for a time , the number of chromosomes is very not the same in different nuclei\mdash ; a fact which is no doubt due to the unequal numbers of constituents taking part in the fusion .
In this division the new cell-wall is laid down in a cell plate in the usual manner ( fig. 36 ) .
As a result of these cell divisions , the lower part of the sac enlarges , passing through the stages outlined in figs. 89 , .
This primary growth , which occurs before fertilisation , ceases at about the stage of .
A more considerable secondary growth commences after fertilisation .
The endosperm-tissue , both primary and secondary , is characterised by a marked irregularity in the size of its cells .
This seems to be due to the fact that when division occurs the daughter-cells are smaller than those from which they originated , and with successive divisions the difference becomes more marked .
Consequently throughout its life the endosperm consists of cells , large and small and of intermediate sizes , intermingrled ( fig. 88 In cases in which fertilisation does not take place , the sterile tissue which fills the lower part of the sac undergoes little further growth , and eventually becomes disorganised .
An embryo-sac-tube has been figured ( fig. 30 ) , in which all the nuclei have fused .
If the product is a fusion-nucleus , homologous with that of a primary endosperm-cell , it should , in the ordinary course , undergo mitosis , and proceed to form a tissue .
A fusion-nucleus in this position has only been seen twice , and it is clearly of rare occurrence .
The ucture shown in fig. is contained within an embryo-sac-tube , and is situated a little below the level of the line in the text-figure .
It can only be an extremely abnormal embryo or a small endosperm tissue formed from such a * STRASBURGER , 1880 , Taf .
3 , figs. 61-64 .
PEARSON , 1906 , p. 266 .
PEARSON , , figs. S PEARSON , , p. 292 .
See PEARSON , 1906 , p. 293 , fig. PROF .
H. H. W. PEARSON : FURTHER OBSERVATIONS ON CHTA .
353 nucleus as those shown in fig. 30 .
Among the hundreds of pro-embryos in various stages that have been examined , the structure and order of development is practically constant .
And since , further , no pollen-tube communicates with the cavity in the nucellar cone in which this ube lies , it may safely be concluded that the cell-mass is not derived from an oospore .
On the other hand , the seven cells of which it is composed are exceedingly like the smaller cells of the normal endosperm of the same sac ( cf. figs. and B ) .
The resemblance is seen not only in the size of the cells and in the highly vacuolated condition of their cytoplasm but in the size of the nuclei , most of which contain more than a single nucleolus .
In the absence of intermediate stages it is impossible to be quite certain that this is an endosperm , but there can be very little doubt that this is its true nature .
Similar structures similarly situated have been seen in two other cases .
The nuclear fusion which precedes the formation of the endosperm occurs in all those compartments which contain many nuclei .
These normally constitute nearly .the whole of the lower three-fourths of the sac .
But among them are very frequently a few in which the nuclei are in smaller number\mdash ; six or less .
In these , the nuclei sometimes , probably always , do not fuse .
These may occur anywhere in the sterile part of the sac , and one or more of them is not uncommonly found close to or at the antipodal end figs. .
Except for their position these compartments differ in no respect from those of the fertile end of the sac , and it is thel.efore not surprising that they should behave in a similar manner .
The compartment enlarges at the expense of the surrounding cells and vacuoles appear in the cytoplasm .
Its wall usually grows upwards , , or vnwards , as a tube eventuaIIy penetrates the nucellus like the normal embryo-sac-tube ( fig. ; text Sometimes when it is deeply immersed in the endospel.m its growth is equal in all directions , and thus no tube , but merely an rged cell , is produced ( fig. When a tube is formed , the free nuclei and cytoplasm move into it in precisely same way as in the normal embryo-sac-tube\mdash ; it is , in fact , an sac-tube remarkable only for its deep-seated origin .
In some cones ahnost every ovule shows one or more of these abnormally situated tubes ; in others they occur in very few .
The endosperm-cells lying near them are frequently bi-nucleate , probably as a result of amitotic division .
It has not been possible to prove that the nuclei of any of these tubes become fertilised , but .
history , behavioul , and appearance leave no room for doubt that they are functionally sexual .
One unpollinated ovule has been found in which no nuclear fusion has occurred , but all the compartments behave like those of the micropylar of a normal sac ; nearly all have produced upgrowing tubes into which their cytoplasm and free nuclei have moved ; a few have not formed tubes , but are clearly quite similar to the rest , and are either less advanced in development or have ) depleted by their more active neighbours .
Most of the tubes have already penetrated the nucellar cone .
* These cells frequently become many times larger than the one figured .
VOL. CC.\mdash ; B , 2 .
H. W. PLARSON : FURTHER RVATIONS ON WELWITSCHIA .
abnormality of this sac extends also to its form .
Its breadth is approximately equal to its length ; in bulk it is about as large as a normal sac in a corresponding .
No more than six nuclei are present in any of the tubes or in the compartments which have not yet formed tubes .
In this case there is nothing to indicate the causes which have induced so unusual a course of development .
In another cone whose ovules have yielded am exceptionally large number of deep-seated sexual cells , the terminal part of the axis , with its bracts and young flowers , is arrested , presumably as the result of injury\mdash ; perhaps caused by insect puncture .
But however these remarkable bnormalities m have been incited , it is of the greatest interest that under certain circumstances all the nuclei the sac may behave like the functional ametes of a normal , at least up to the stage of fertilisation .
And since the their nuclei are all precisely alike\mdash ; excepting those few whose contents have \mdash ; both in appearance and reactions , and are further quite similar in these respects to the tubes and nuclei which in normal ovules become fertilised , it cannot doubted that these also are potentially sexual .
Another abnormal sac presents a case almost the converse of the last , and the two considered together with the occasional production of a fusion-nucleus within the embryo-sac-tube and the frequent occurrence of compartments con- taining sexual nuclei among the endosperm-cells leave little doubt as to the potential of the sac-nuclei .
In this instance the ovule contains two sacs , one above other , the contiguous surfaces being closely pressed together .
In the upper one septation has just commenced ( fig. 39 ) and all the nuclei are free ; the lower is considerably more advanced and the endosperm is constituted each of its cells containing a spherical fusion-nucleus .
normal sac with endosperm-nuclei in this condition would show embryo-sac-tubes as advanced as those of . .
It is extremely interesting to find that where the top of the sac in this case is in contact with the lower wall of the one above it , not only are there no embryo-sac-tubes , but all the nuclei have fused , and the compartments which under ordinary circumstances would have contained free sexual nuclei have become uninucleate endosperm-cells .
The significance of this receives emphasis from the fact a few peripheral compartments of the micropylar region of this sac ( which being ) advanced is broader than the upper one ) still retain their free nuclei and some of are in an early stage of tube-formation ( fig. 39 , .
In otl , er words , where the micropylar end of the lower sac is subjected to a vertical pressure , the rtments have become endosperm-cells by the fusion of their nuclei ; where , as at the periphery of this region , there is little resistance to upward growth , this has occurred , and the nuclei are still free and have all the characters of the functionally sexual nuclei of a normal sac .
This is arded as confirming the opinion already expressed ( p. 348 ) , that the less crowded condition of the micropylar nuclei which normally precedes septation ( fig. 22 ) is due to the more active growth in length of this end of the sac .
When this is PROF .
H. H. W. PEARSON : FURTHER OBSERVATIONS ON LWITSCHIA .
355 prevented , as in this case by a superincumbent sac , the nuclei do not separate and the compartments consequently contain several nuclei instead of a few , and fusion occurs .
The various abnormalities , which have now been described are only explicable on the assumption that the free nuclei in the sac at the time of septation are all alike in every respect\mdash ; all potential gametes .
And since both the history of their origin and their visible characters are also in favour of this conclusion , it may be regarded as proved .
The lower part of the sac owes its sterility solely to the fusion of its nuclei , and this is confined to those of its compartments in which fusion has occurred .
An endosperm*formed in this way is clearly not a prothallus , for it is formed as a result of the fusion of sexual nuclei ; it is consequently not a part of the gametophyte generation ; and further , it does not produce sexual organs or sexual cells .
Neither is it a part of the sporophyte generation , for its nuclei contain more than the prelneiotic number of chromosomes and it p.lays no direct part in the production of sporogenous tissue .
It possesses a very considerable power of primary and secondary growth ; the latter is not manifested uniformly throughout its bulk , but , as will be shown later , is to some extent localised .
The endosperm of Welwitschia is therefore an organism showing a degree of specialisation .
It is not represented in the lower Gymnosperms or in the Pteridophytes .
It is not , however , a newly intercalated g.eneration , for it does not stand in the direct line of the life-cycle of the plant .
It is rather a bye-product resulting from the fusion of potentially sexual nuclei , and functioning in the same manner as the prothallus of the lower seed-plants which it replaces in the nutrition of the embryo .
There is reason to believe that an endosperm of this character is of general occurrence in the higher seed-plants , and it is necessary to distinguish it from the prothallus .
It is therefore proposed to designate it by the term Trophophyte .
Its relation to the other stages in the life-history may be thus represented:\mdash ; Sporophyte .
Sporophyte .
Microspore .
Macrospore .
Gametophyte reduced to one evanescent vegetative nucleus , a tube-nucleus , and a generative cell .
Gametophyte , a plasmodium Male nuclei .
containing free female nuclei .
mbryo .
Trophophyte .
* Endosperm is a physiological term ( RICHARD , 1825 , p. 353 ; LINDLEY , 1848 , vol. 2 , p. 55 ) .
It is eqwlly applicable to the prothallus of the gymnosperm as to any form of nutrive tissue ised within the embryo-sac of a seed-bearing plant .
PROF .
H. H. W. PEARSON : OBSERVATIONS ON WELWITSCHIA .
If under any circumstance two female nuclei only should fuse , the fusion-nucleus would possess the number of chromosomes characteristic of the sporophyte and a normal elnbryo m perhaps result\mdash ; as when two post-meiotic nuclei of vegetative cells fuse in the prothallus of Lastrea , WILLS .
* Evidence that this occurs in Welwitschia has been searched for in vain .
In view of the similar conditions o.f the Gnetum sac\mdash ; to be discussed later\mdash ; it is possible that the supposed parthenogenetic embryo in G. may originate in the fusion of two of the nuclei .
A comparison of the endosperm of Welwitschia as regards its manner of origin with that of Ephedra on the one hand and of Gnetum on the other leads to important conclusions .
here can be no doubt that in Ephedra the endosperm is a true prothallus formed in the same way as that of the Conifers , the Cycads , and the higher Pteridophytes .
Its development has been studied in four species and , in the free-nuclear embryo-sac becomes septate as a result of the formation of centripetal tubular alveoli .
In the later stages of its development multinucleate cells are met with , as in some other Gymnosperms .
In Taxus , many of the endosperm-cells , : all of which are initially uninucleate , become multinucleate\mdash ; whether by direct or indirect division is not certain .
These cells may eventually contain as many as 16 nuclei .
After this nuclear fusion occurs , and the number of nuclei in each cell is , in consequence , reduced .
This is succeeded by their degeneration .
JAEGER observes that the degeneration occurs first in those endosperm cells which border upon the archegonia or embryo .
Whatever may be the significance of the multiplication followed by fusion and degeneration of the nuclei of endosperm-cells , it is clearly a process in no way connected with the nuclear fusion which produces the Welwitschia endosperm , and its occurrence does not lessen the fundamental difference which exists between the prothallus of Ephedra the trophophyte of Welwitsehia .
At the micropylar end of the prothallus of Ephedra , as of the lower Gymnosperms , there arise sexual organs whose homology with the archegonia of the Pteridophytes cannot be questioned .
Welwitschia and Ephedra are therefore very widely separated by the character of the endosperm and of the sexual apparatus .
It is unfortunate that the development and the minute structure of the endosperm of Gnetum has been but little investigated .
The species that have been studied fall into groups .
In the one which appears to be most numerously represented , * FARMER and DIGBY , 1907 .
LOTSY , 1903 .
, 1872 , Taf .
16 , fig. 52 ; SOKOLOWA , 1891 , fig. 22 and .
; JACCARD , 1894 , p. 17 ; BERRIDGE and , 1907 , p. 129 .
S JACCARD , .
cit. , fig. 48 ; STRASBURGER , 1880 , p. 106 .
JAEGER , 1899 , pp. 258 and , Taf .
17 , fig. 29 ; STRASBURGER , 1880 , p. 107 ; LAWSON , 1904 , p. 426 .
JAEGER , , and Iig .
30 .
KARSTEN , 1892 , 1893 .
PROF .
H. H. W. PEARSON : FURTHER OBSERyATIONS ON WELWITSCHIA .
357 sperm is not formed before fertilisation .
The embryo-sac becomes filled with free nuclei embedded in cytoplasm , as far as can be judged , in the same way as in Welwitschia , except that in the early stages there is normally a large central vacuole .
All the nuclei after the last , divisions have occurred are said to be alike in and potentially sexual in function .
Some of these are fertilised .
As soon as this has occurred the lower part of the sac becomes septate .
According to KARSTEN'S figure of G. edule* the compartments are very irregular in form , and most of them are multinucleate .
In G. lifolium " " in vielen Endospermzellen sieht man zwei Kerne ; es kommen auch haufig weit mehr als zwei auf eine Endospermzelle bei der ersten Zerlegung des Plasmabelages.\ldquo ; It is implied that these compartments later become uninucleate cells and afterwards undergo division .
Nuclear fusion is not recorded , but if it does not occur these species of Gnetum must show still a third method of endosperm-formation differing alike from those which are characteristic of Ephedra and Welwitschia respectively .
This is exceedingly improbable .
Gnetum Gnemon must be separated from the species studied by KARSTEN , on the ground that an Josperm is formed , at least in many cases , before fertilisation .
There is every indication that at the end of the free-nuclear condition of the sac the nuclei are equivalent in every respect .
Those of the lower part of the sac form a sterile tissue , which later fills the sac and nourishes the pro-embryos , while any of those in the upper part may become fertilised .
LoTSY gives little detail as to the manner in which this sterile tissue is formed , but the followin sentence is significant : \mdash ; " " We notice that in fig. 29 , Plate 4 , several cells contain as yet a number of nuclei , while in fig. 34 , Plate 5 , this has already been remedied , eyery cell containing but one.\ldquo ; S The nature of the remedy has not been ascertained , but in view of what occurs in Welwitschia it is reasonable to conclude that here also it is nuclear fusion .
After fertilisation " " a greater or smaller number of the nuclei of the fertile part of the embryo-sac surround themselves with a denser protoplasm , a membrane , and thus form cells These " " retarded prothallium cells\ldquo ; do not persist , being eventually obliterated and replaced by the endosperm growing up from below .
more active growth of the latter tissue is no doubt he result of the stimulus due to the fusion of its nuclei .
One other feature of the endosperm of G. described by Lorsy possibly has a bearing on this discussion .
There occur occasionally at the summit of the endosperm groups of cells of a remarkable form , " " which one is compelled to consider as rudimentary archegonia.\ldquo ; One of these is shown in section in fig. 62 .
One of , 1892 , fig. KARSTEN , 1893 , p. 338 .
KARSTEN , .
cit. , p. 372 .
S LOTSY , 1899 , p. 98 .
cit. , p. 97 . .
cit. , p. 98 .
: it 58 PROF .
H. H. W. PEARSON : FURTHER OBSERVATIONS ON WELWITSCHIA .
the cells of the group is much larger than the rest , and contains several nuclei .
Its general similarity to one of the compartments of the endosperm of Welwitschia in which nuclear fusion has not occurred is certainly striking .
It must be admitted , however , that this interpretation is not confirmed by other figures of similar structures cf. figs. 55 and 56 though it is not clear that even in these cases the cells are originally uninucleate .
It is a reasonable inference that in the species in which it is formed after fertilisation , as well as in , in which its organisation is usually delayed , the of Gnetum is a trophophyte essentially to that of Welwitschia .
G. ermination of the Pollen-grain and Gvowth of ' the Pollen-tube ( Plates 26 and 27 , figs. 40-52 ) .
exine splits into two lobes , and may be throwll off entirely as in Ephedra .
The tube-nucleus passes into the growing intine in advance of the generative cell .
latter increases considerably in size during the early stages of germination figs. , and lies quite free in the cytoplasm of the pollen-grain .
As it leaves the body of the grain it becomes elongated in the direction of the axis of the tube and narrower in proportion .
The sbape of its nucleus is usually affected in the same way ( fig. 41 ) .
The nnterior end of the cell is bluntly pointed , while posteriorly it is thrown out into a number of irregular lobes .
Both these characteristics are usually found up to the time of fertilisation .
The nucleus possesses a single large nucleolus .
The pollen-tube usually enters the nucellus at the summit !
sometimes it continues to grow in the micropylar fluid between the integument and the side of the nucellus ; it occasionally passes beyond the line in the text-figure before turning sharply lwards and penetrating the nucellus .
Soon after entering the nucellus both the generative cell and its nucleus a further increase in size ( fig. 43 ) .
In front of the cell there is usually a vacuolar space .
The tube-nucleus frequently lies , in these early stages , about midway between the tip of the tube and the front end of the generative cell ( figs. and 46 ) ; it is often , however , almost in contact with the latter .
Its form is somewhat irregular , but it is most frequently more or less elongated and crescent-shaped ( figs. 43 , 45 , 46 ) .
In the young tube it is usually sharply differentiated from the surrounding cytoplasm , but as it approaches the level of the line ( text-figure ) its outline becomes indistinct , and below this line the signs of its disintegration are always well marked ( fig. 50 ) .
Its nucleolus , which is frequently as large as that of the male nucleus , is the most persistent part of it .
At the time of fertilisation even this is rarely recognisable .
, 1899 , .
98 .
PEARSON , 1906 , fig. 17 .
, ls94 .
S , 1906 , PROF .
H. H. .
PEARSON : FURTHER TIONS ON WELWITSCHIA .
359 The identification of the male and female nuclei just prior to the process of fertilisation is often difficult , even when the latter is quiescent in a large oosphere .
Here the lnale nucleus is easily distinguished by its cytoplasmic sheath , its constantly large nucleolus and , very frequently , by its elongated form .
More serious difficulty is occasioned by the resemblance between some forms of the tube-nucleus and the female gametes .
In view of the fact that fertilisation occurs within the male cytoplasm , it becomes especially important to distinguish the one from the other .
In -fixed material the female 1lucleus is approximately spherical in form and its nucleolus is almost invariably smaller than that of the tube-nucleus .
More important , however , is the fact that before contact between two tubes is effected the tube-nucleus is either already absorbed or at least shows distinct signs of deterioration .
Occasionally , however , undoubted female nuclei have been seen which very closely resemble a common form of the tube-nucleus ( figs. 29 and 46 ) .
While these are ahnost certainly iorated in fixation .
occasional occurrence renders necessary the exercise of considerable caution in identifying female a nucleus lying just in front of the male cytoplasm after communication between the two tubes has been established .
While the direction of growth of the -sac-tubes is in general more or less away from the axial line of the nucellar cone , that of the pollen-tubes is rather towards it .
The meeting of the two sets of tubes occurs usually near the axis of that part of the nucellar cone which lies between the lines and ( text-figure ) .
Fertilisation may be effected , however , in any part of the nucellus lying above the top of the embryo-sac .
Pollen-tubes which do not come into with enlbryosac-tubes in the nucellar cone still continue their downward growtI ] .
Insffinces of this are not common , for the lower part of the nucellar cone is thoroughly perforated by embryo-sac-tubes , always more numerous than the )-tubes , their is almost ensured ; added to there is also the probability that they are definitely attracted towards one another by chemiotaxis ( fig. 53 ) .
Nevertheless a pollen-tube will rarely find its way down into the evacuated fertile rion of the sac almost to the top of the endosperm itself , where its contents are no absorbed by the actively growing any case the nuclei of such a tube can exert no influence upon the growth of the endosperm ; apart from the very small number of cases in which ] they are brought into its vicinity , the primary -nuclei are fully formed and cell-division has always collmenced before pollination occurs .
In one ovule which contains a number of embryos , a pollen- tube , grown through the region occupied by the -tubes , has reached a point in the chalazal tissue about half-way between the base of the endosperm and the insertion of the ) .
Its course lies entirely the nucellar tissue ; it is not shown in one section but is indicated by dotted lines in fig. 47 .
Near its tip is the bi-nucleate generative cell figs. and ; the tube-nucleus has disappeared .
PROF .
H. H. W. ARSON : FURTHER OBSERVATIONS ON WELWITSCHIA .
At the time of pollination the cells of the nucellar cone are packed with spherical starch-grains which stain readily in diamant fuchsin .
Tn this condition they are seen to be composed of three or four simple grains , each of which has in section the of sector of a circle and shows a minute central hilum ( fig. 42 ) .
Their compound structure is not brought out by any of the other stains used and they were formerly described as simple and concentric .
* The starch-grains of the endosperm have the same structure .
Soon after entering the nucellar cone the nucleus of the generative cell divides ( figs. 44 and 45 ) .
The spindle is indistinct ; its axis is parallel with that of the cell .
The two nuclei resulting from this mitosis are at first spherical in form and lie near together at either end or in middle of the cell ( figs. 46 , 49 , 50 , etc They undergo further division ; both are potentially sexual .
rrhc ucleate generative cell shows a considerable range of differences in its later behaviour .
The nuclei very frequently exhibit a tendency to elongate ( figs. 51 , 54 , 55 , etc which may become verypronounced ( figs. 59 and 60 ) .
Whatever their the leading nucleus is usually smaller than the one behind it ( figs. 46 , 48 , 51 , etc Sometimes both the time of fertilisation ( fig. 57 ) ; when the cytoplasm of the cell remains intact one of the two nuclei frequently\mdash ; perhaps usually\mdash ; breaks down ; may be either the anterior ( fig. 52 ) or the posterior ( fig. 59 ) .
Sometimes the cytoplasm of the cell becomes separated into anterior and posterior portions , each containing one .
It is probable that this , when it does occur , is a purely mechanical effect of the strains caused by the passage of the cell through a narrow and commonly tortuous passage .
Since these ucleate masses of cytoplasm are not homologous with the male cells of the lower Gymnosperms , the term ' ' generative cell\ldquo ; is used indiiferently for the cytoplasmic sheath , broken or entire , enclosing , in the former case , one nucleus , and in the latter , one or two nuclei .
At the time of fertilisation , then , the enerative cell may contain either two male nuclei or a single male nucleus , the other having disintegrated , or , on the other hand , it may broken into two parts , each containing a single potentially sexual nucleus .
The male nucleus is characterised by a Slngle nucleolus which is always distinctly larger than that of the female .
In most cases there is a considerable quantity of extrachromatin .
-nuclear granules staining deeply in diamant fuchsin are also usually present .
In no case does the nuclear body show any signs of a spiral curvature .
The development of the male gametes from the generative cell differs remarkably from the corresponding series of events in Ephedra and the lower Gymnosperms ; it appears , however , to be quite similar to those which occur in some species of Gnetum .
In the omission of the division which cuts off the ' stalk- " " or " " wall-cell " " and in the fact that the male luclei are free within the generatiVe cell , Welwitschia and * PEARSON , 1906 , p. 292 .
KARSTEN , 1893 , figs. 59-64 .
PROF .
H. H. W. PEARSON : FURTHER OBSERVATIONS ON WELWITSCHIA .
.3(i1 Gnetum show a close approach to the condition of many Angiosperms .
* The last gymnospermic feature ined in the structure of the pollen-grain is the reduced and evanescent prothallial cell , such as is also described for a few Angiosperms .
H. FertilisrPlates figs. .
While in the early stages of the pollen-tuk ) the tube-nucleus is usually more or less in advance of the generative cell and some distance ) the tip of the tube ( fig. 46 ) , when the latter crosses the line in the text-figure they lie near together and not far from the apex ( fig. 50 ) .
By this time the tube-nucleus has lost the greater part of its chromatin and its outline is commonly indistinct .
As the tube advances , the body of the nucleus becomes less distinct and finally the nucleolus disappears .
The generative cell , or the leading portion of it , then comes to lie just within the tip of the pollen-tube ( figs. 51 and 55 ) or a very little way behind it ( fig. 63 ) .
As the embryo-sac-tube approaches the pollen-tube its leading nucleus is almost invariably considerably in advance of the rest .
It is on this account and owing to the complex of tubes perforating the lower part of the nucellar cone that it is frequently impossible to identify the other nuclei of a tube whose leading one has become fertilised .
After its fertilisation it seems that the others of the same tube degenerate ( figs. 65 and 67 ) .
The fertilisation of more than one of the female gametes of a tube is theoretically , but no instance in which it has certainly occurred has been observed .
A possible case is that of the tube whose upper end is shown in fig. and which is continuous below with a cavity in the llucellus in which lies a two-celled pro-embryo .
If the nucleus " " 9 ?
\ldquo ; in this figure is correctly tified as a female gamete , is probable that the pro-embryo results from the fertilisation of a second female nucleus in the same tube .
Before the nucellar tissue which them has become disorganised , a pair of tubes are often to be growing very definitely towards one another ( fig. 53 ) , which it may be ferred that they both or one of them is u1lder the influence of a directive force .
Since the most active growth of the embryo-sac-tube occurs before pollination it might be expected that the attractive influe1lce emanates from it .
On the other hand , several cases have been seen in which the tip of the embryo-sac-tube comes into contact with the -tube , not at its apex but laterally in the neighbour- hood of the generative cell and 63 ; this suggests that the growth of the embryo-sac-tube is rather directed towards that part of the pollen-tube in which the generative cell lies .
The walls at the place of contact break down and their conteIlts are placed in communication .
The leading female nucleus passes directly into the generative cell within which fertilisation takes place .
This very remarkable behaviour COULTER and CHAMBERLAIN , 1903 , pp. 133 and , 1906 .
It is very improbable that this is an unusually persistent tube-nucleus , bnt this is caso which its identity is somewhat doubtful .
VOL. CC.\mdash ; B. 3 ,362 PROP .
H. H. W. OBSERyATIONS ON WELWITSCHIA .
of the female gamete been carefully ascertained and there can be no doubt that it is normally antecedent to the nuclear fusion .
Fig. shows both the male nuclei , the generative cell having broken in the middle ; within the cytopl of the leading ) ortio of it , lying immediately in front of its nucleus , is an undoubted lllcleus .
this case the lower part of the embryosac-tube comes immediately from a complex of similar tubes and its other nuclei cannot be certainly tified .
The male llucleus of the hinder part of the generative cell shows sigllS of ient degeneration .
A similar case is shown in figs. ; here , in three of four successive sections , are shown : ( 1 ) male and a female nucleus within the generative cell ( A ) ; ( 2 ) the same male and a second female nucleus behind the first one and not within the generative cell ( B ) ; ( 3 ) a third female nucleus considerably lower down in the same embryo-sac-tube ( C ) .
The posterior male nucleus disappeared .
-sac-and pollen-tubes in communication are shown in two successive sections in figs. 62 and 63 .
In the former a female nucleus is in contact with the nlale within the generative cell ; in the latter there is a second female ucleus at a lower level in the same embryo-sac-tube .
Other cases in which the female nucleus is clearly the geIlerative cell are shown in figs. 61 and 64 .
In the latter the second male nucleus is not to be found ; in the former it is probably represented by of chromatin in the hinder of the generative cell .
The cytoplasm of the generative cell is the most conspicuous structure within the -tube ; the general cytoph ) of the tube as well as all that of the embryo-sactube is , when fertilisation occurs , uniformly distributed and vacuolated .
If , then , the fusion ths bamete were ever effected outside the limits of the generative cell , the remains of the latter would certainly be visible for a time in the neighbourhood of the oospore .
A many } ) ores have been examined , but such an extraneous mass of cytoplasm has never been fOund in the neighbourhood of any of them except in a few cases in which it contained male nucleus and was clearly the second half of the enerative cell .
The cytoplas1n of the oospore is so massive ( figs. 65-67 ) and that 01 the embryo-sac-tube so lacking in organisation , that the former must be derived very largely , if not entirely , the generative cell .
Since in the young oospore the cytoplasm frequently does not stain uniformly ( fig. 65 ) , it is probable that some nies the female nucleus from the embryo-sac-tube .
But there is no doubt that in Welwitschia the male cytoplasm plays a much more prominent part in the process of fertilisation than ] been found to be the case in other Gymnosperms .
* In some cases , at least ( c.g. , figs. 57 , 63 , 64 ) , the generative cell certainly does not leave the pollen-tube , but the female nucleus passes out of the embryo-sac-tube to enter it .
latter is identified in each case by the presence within it of a second female nucleus below itsjunction with the pollen-tube .
Owing , however , to the large number of tubes and their very devious courses , it is usually impossible to determine with certainty in which of the two tubes the generative cell at the time of fertilisation , COKER , 1907 , p. 6 .
PROF .
H. H. W. PEARSON : FURTHER OBSERVATIONS ON WELWITSCHIA .
363 or the oospore , lies .
But no case has been found in which fertilisation clearly occurred within the embryo-sac-tube and it is that it always takes place while the generative cell lies within the pollen-tube .
This would account for the connection which HOOKER cDnstantly found between the pollen-tube and the primary suspensor of the pro-embryo .
* As the male and female nuclei come into contact ( figs. 61 and63 ) , a concavity frequently appears in the former .
Sometimes howevel it seems to be the male nucleus which impresses the surface of the female .
The fusion ( fig. 64 ) is rapidly completed , but the nucleoli remain distinct for a time ( figs. 65 and 66 ) .
They are still to be distinguished by their difference in size .
At first the body of the nucleus is differentiated into two parts , of which one stains more deeply than the other ( fig. 65 ) ; in later stages this difference has disappeared ( figs. 66 and 67 ) .
A cell-wall is formedjust within the limits of the generative cell .
In some cases a second female nucleus , in a condition of degeneration , is included within the delimited cytoplasm ( fig. 65 ) .
A resting nucleus is thus formed as a result of fertilisation and in the first mitosis the paternal and maternal chromosomes are not to be distinguished ( figs. 68 and 69 ) .
This is the case also in Lilium and other Angiosperms , while in Pinus and perhaps in Gymnosperms generally , S a resting fusion-nucleus is not formed .
account of Ephedr trifurca suggests that Ephedra and Welwitschia are in agreement in this respect , but the first mitosis in the oospore of the former has not been described .
The details of fertilisation are insufficiently ascertained in Gnetum , in which fusion of the male and female nuclei has not been observed .
In the male nuclei are free in the cytoplasm of the pollen-tube , and in this condition enter the embryo-sac .
In other species , however , the male gametes are surrounded by the well-defined cytoplasm of a generative cell ; they are discharged into the embryo-sac , each surl.ounded by \mdash ; presumably that of the gelJerative cell .
The nuclei are somewhat elongated and narrowed towards the anterior end , as is so frequent in the case of Welwitschia .
Whether the cytoplasm accompanyin them takes part in the organisation of the oospore was not determined , though formerly inclined to the opinion that this was the case .
SS , 1863 , , figs. MOTTIER , 1904 , p. 176 . .
S DAVIS , 1905 , p. 232 .
LAND , 1907 .
LOTSY , 1899 , p. 94 .
, 1893 , fig. 65 , etc. ; LoTSY , KARSTEN , loc. , p. 379 .
KARSTEN , loc. ( it .
, fig. 71 .
S KARSTEN , 1892 , p. 224 .
I. The Plates 29 and 30 , figs. 68-87 number of oospores present in a nucellus is very variable .
It is rarely so few as one , and probably never exceeds the number of the pollen-tubes .
It is not uncommon to find as many as a dozen pro-embryos in an ovule .
After the formation of the single nucleolus , the oospore and its nucleus increase in size and become elongated in the directiori of the axis of the pollen-tube within which it lies .
The boundary of the oospore-cytoplasm approximately coincides with that of the original generative cell ( fig. 67 ) .
Where , however , both the male nuclei have persisted , only the anterior part of the generative cell contributes cytoplasm to the oospore .
This is probably the reason of some differences in size frequently exhibited by oospores in the same nucellus .
When the wall is completed the elongation of the cell becomes more pronounced and vacuoles appear in the cytoplasm on the micropylar side of the nucleus ( figs. 68 and 69 ) .
This growth is certainly for the most part in a downward direction .
But the frequent occurrence of an irregular projection at the micropylar end probably indicates that at first the cell elongates in both directions .
Since , in later stages , the micropylar end of the pro-embryo is always found in contact with unbroken nucellar cells , it seems that the upward growth continues until a point is reached .
A very similar projection at the micropylar end of the elongating oospore of Gnemon is described and figured by LoTSY .
The cells at first lie entirely within the canal excavated by the pollen-tube and embryo-sac-tube .
Iu many cases the pro-embryo continues its course down the canal formed by the embryo-sac-tube , which formerly brought the female nucleus which has given rise to it .
On the other hand , it frequently bores a new path through the hitherto uninjured nucellar tissue : Gravity can have little if any influence upon its direction of growth .
Apart from the fact that the female cones , and with them the ovules , are usually inclined at a considerable angle to the vertical , there are circumstances in which the pro-embryo will grow directly upwards towards the micropyle .
This has been observed only in a few cases , in all of which fertilisation has occurl.ed above the line ( text-figure ) .
It would seem therefore that the direction of growth is determined mainly by conditions of nutrition and , further , the downward course will usually be that of least resistance .
It may be that oospores situated near the micropyle are within the range of influence of the sugar left by the drying up of the micropylar fluid .
As the cell elongates the nucleus moves downwards and , when not far from the lower end of the tube , undergoes mitosis ( figs. 68 and 69 ) .
At the conclusion of this division a wall is formed , transverse to the axis , dividing the oospore into two cells , of which the lower one is much smaller than the upper ( iig .
71 ) .
The origin of this of the stages ) elow : figured in outline by . .
This sometimes occurs netum oTSY , .
PROF .
H. H. W. iSON : ] , ON transverse wall ] not been certainly ined .
nrin the mitosis ( fig. is no visible cell-plate .
a case in which the of Ilsverse wall is very distinct while in the , between Tcxt-figure.\mdash ; A longitudiu section , of ovules , one at the of tion ( on the right ) , the other ining a ) ryos ( 011 tho loft ) .
The is more highly gnified than the length .
cavity in endosperm caused by the entrance of ) ryos ; ) primary : seco1td outer(epithelitl ) layer of cone figs. ) ) ; integument ; micropyle ; nucellus , oosporo ; -celled ( .
fig. 73 ) ; -celled ) ( .
fig. 76 ) ; later ) enlI ) fig. pro-embryo cf. fig. 85 ) ; pollen-grain ; polle1l-tnbe ; .
of ondospcrm ; egion of the endosperm which the cells moro is presc.nt clsewherc .
the nuclei , it cannot be traced .
It is therefore probable that , as in Ephoedra trifurca , arises as an ingrowing ring from the wall of the oospore .
The upper of the two cells of which the pro-embryo now consists undergoes no division ; it attains a considerable length as the " " primary\ldquo ; suspensor and later as the axial part of the " " secondary\ldquo ; suspensor .
The lower cell quickly divides , to give rise to a group of four initial cells from which are later formed : ( c ) twenty-four " " cortical\ldquo ; cells of the secondary suspensor ; ( b ) the embryonic plate of eight cells ; ( c ) a ring of sixteen cells around the latter ; ( d ) a terminal ' ' cap\ldquo ; of cells .
As the primary suspensor elongates , pushing the lower cell down in front of it , its cytoplasm becomes vacuolated and its nucleus maintains a position not far its lower end ( figs. 71 and 72 ) .
The terminal cell now divides longitudinally ( fig. 7 and the two cells ( fig. 73 ) divide again in a plane at right angles to that of the first division .
At the lower end of the primary suspensor there is now a group of four beak-like initial cells ( fig. 74 , Each of these divides in a plane at right angles to the two former divisions .
One cell undergoing this division is shown in fig. 75 ; the other three , in practically the same phase , lie in the next section , but are somewhat displaced in preparation .
There results a group of eight cells in tiers ; those of the upper tier are smaller than those of the lower ( fig. 76 ) .
The former divide by radial walls parallel to the axis of the pro-embryo ; this division is in progress in fig. 77 .
The upper tier now has its full complement of eight cells ; it later gives to the inner cortical ring of the secondary suspensor ; it is therefore conveniently referred to as the " " inner cortical tier\ldquo ; .
By this time the elongation of the primary suspensor has brought the terminal group of cells into the region formerly occupied by the fertile end of the embryo-sac ( text-fig .
) .
four terminal initial cells divide by radial longitudinal walls meeting in the axis of pro-embryo .
Henceforward the initial group contains eight equal cells .
Each of these cuts off a cell by an oblique wall ( fig. 78 , A and B ) .
The cells so separated form a ring lying above and outside the initial cells and beneath the inner cortical tier ( fig. 79 ) .
The eight cells of this ring further divide by radial longitudinal walls producing the sixteen initial cells of the outer cortical ring figs. .
The next stage is marked by the further division of the eight terminal initial cells by walls transverse to the axis of the pro-embryo ( fig. 80 ) .
There is thus formed a plate of eight cells figs. , which lie within the outer cortical ring and form what is provisionally called the " " embryonic plate\ldquo ; ( E ) .
When this stage is reached the terminal cell-group of the pro-embryo has usually penetrated the upper part of the endosperm and the formation of the secondary suspensor commences .
The first sign of this is the elongation of the *LAND , 1907 , figs. 14 and 15 .
PROF .
H. H. W. PEARSON : FURTHER RVATIONS ON WITSCHIA .
367 eight cells forming the inner cortical tier ( i.c.t. ) , which has already begun in fig. 80 .
The direction of growth of these cells is that of the elongation of the whole pro-elnbryo .
The initial stage of their elongation is marked by the appearance of a vacuole in the proximal part of the cytoplasm .
The growing region is clearly intercalary and is situated at the angle where the cell curves in under the primary suspensor .
In view of HOOKER'S figures , is to be emphasised that these eight cells together with the distal end of the suspensor grow on together as a tissueunit ; they are not isolated " " embryonal tubes\ldquo ; with independent growth in a backward direction as has been believed .
The same is true of the cells of the outer cortical ring .
While the eight cells of the inner cortical ring are in an early stage of elongation figs. , the nucleus of the primary suspensor-cell shows signs of disintegration and in later stages cannot be distinguished ( fig. 82 ) .
Meanwhile the eight terminal initial cells undergo a division similar to that which formerly cut off the outer cortical ring .
A ring of cells , afterwards by radial longitudinal division increased to 16 , is thus formed lying above and outside the initial cells and beneath the outer cortical ring .
These 16 cells are unchanged in the most advanced embryo yet found .
One of HOOKER'S figures ( Plate 10 , fig. 25 ) suggests that t may later giye rise to a third ring of cortical suspensor cells .
They will be referred to here as the " " lower ring\ldquo ; ( figs. 81-85 , , " " The 16 cells of the outer cortical ring now begin to elongate in precisely the same manner as did those of the inner ring .
Fig. 84 , section , shows four cells of the inner cortical ring , two of the outer , one ich may belong to either , five embryonic cells , two cells of the " " lower\ldquo ; ring , and four of the terminal cells .
The outer cortical cells are considerably more elongated in fig. 85 , but no further change has occurred .
This figure represents the most advanced pro-embryo seen .
It recalls that of Until it is possible to follow its development further it will be best to assulne that terlninal group of eight cells formed when the lower ring-cells cut off the initial cells ( figs. 81 , etc. , " " is homologous with the " " embryo-cap\ldquo ; of Araucaria ; like latter , it is certainly absorptive in function ( see fig. 85 ) .
It further affords mechanical protection to the plate lying above it , which , from analogy with Araucaria , is regarded as the initial embryonic group The relations of the cells composing the pro-embryo of the of fig. 85 will be made clear by reference to the series of diagrams of fig. 86 and to esenting transverse sections at higher elevations .
Fig. is taken from * HOOKER , 1863 , Plate 10 .
( These figures were ined from ) ) rescrved m p. 36 .
) STRASBURGER , 1879 , Taf .
20 , figs. 66-71 .
More advanced stages are figured by ( 1879 , figs. 87-93 ) ; but the devclopment of this plate is obscure ( E ) .
PROF .
H. H. W. : FURTHER RVATIONS ON WELWITSCHIA .
a section at the level of the line marked " " A.\ldquo ; in fig. 85 .
It shows the eight capcells .
A section of\ldquo ; \ldquo ; shows , internally , the eight cells of the embryonic plate ( E ) are surrounded by the 16 cells of the lower ring The section of fig. passes through the cells of both outer and inner cortical tiers below the end of the primary suspensor ( fig. 85 , at the evel " " In fig. 87 is shown a transverse section through the secondary susl ) ltsor which , in the most advanced stage seen , always consists of 25 cells , , the primary suspensor surrounded by eight inner cortical cells outside which are 16 outer cortical cells .
The secondary suspensor is circular in section , the terlninal cell-group is more or less elliptic .
The form of the latter does not appear to be influenced by the flattening of the endosperm , which is usually well marked by the time that it is penetrated by the pro-embryos .
When only one pro-embryo is present it seems usually to pursue a fairly direct course towards the centre of the endosperm .
When , however , as more commonly occurs , several pro-embryos are formed , the core of the upper third of the endosperm becomes totally broken down ( text-fig .
C ) .
In the cavity thus produced by the activity of their own absorptive powers the growing ends of the pro-embryos pursue a somewhat spiral course and a section frequently shows the broken-down tissue of the almost replaced by a most confused arrangement of suspensors cut in all possible planes .
* As a rule one outpaces the rest and thrusts its terminal group of cells into the heart of the endosperm .
The mature seed apparently never contains more than one embryo , and it may be presumed that the pro-embryonal group which first reaches the solid endosperm is destined to survive .
It is now immersed in a tissue whose cells contain a greater accumulation of starch than any other part of the endosperm .
This region is not so markedly differentiated until one or more embryos have entered the endosperm .
The convolutions of the suspensor are so numerous that only short lengths of it are obtained in any one section .
It is therefore very difficult to determine whether transverse walls are formed in the cells composing it .
Both in the primary suspensor and in the cortical cells the nucleus is always found not far from the pro-embryonal group ( figs. 82 and 85 ) ; it has never been seen in division and more than one has not been found in the same cell .
What at first sight appear to be oblique walls occur only when the suspensor has suddenly changed its direction .
It may therefore be inferred that llormally the cells of the suspensor do not divide by transverse walls , but the possibility that these may occasionally occur , as in Gnetum , is not precluded .
The upper ( older ) parts of these without cytoplasm ; their walls react strongly to cellulose stains and show no trace of cutinisation .
* The cavity " " filled with a coiled mass of tubular structures\ldquo ; found at the apex of the endosperm of Gnctnm ( BOWER , 1882 , p. 278 ) is 110 doubt of a nature .
BOWER , 1882 , p. 279 ; LOTSY , 1899 , Plate 10 , fig. 57 .
PROF .
H. H. W. PEARSON : FURTHER OBSERVATIONS ON WELWITSCHIA .
369 The later stages of the Welwitschia embryo and the details of the whole development of the pro-embryo of Gnetum must be investigated before any close comparison between the embryonic stages of the three genera can be instituted .
But in so far as the facts are known , they indicate a closer degree of affinity tween Welwitschia and Gnetum an between either of these genera and Ephedra .
In Ephedra , the oospore-nucleus undergoes three series of divisions , and each of the eight resulting nuclei may organise a cell which gives rise to a pro-embryo .
In some species of Gnetum , the oospore is said to behave in a somewhat similar manner .
* In G. ( inemon and some other species and Welwitschia the oospore produces a single pro-embryo .
In and in E. the formation of the primary suspensor corresponds fairly closely with that of Welwitschia .
What LAND calls the " " secondary\ldquo ; suspensor of E. is , however , very different from that described above is apparently of the same nature as the Cycadean suspensor .
In , the intraseminal development of the embryo is completed while the ovule is still attached to the parent plant .
Indeed , under favourable conditions , germination may occur before the strobilus falls .
In Gnetum , on the other , the development of the embryo itself does not commence until after the seed has fallen .
S None of the cones of Welwitschia that have been examined have ined their full size , but the most fully developed pro-embryos seen ( fig. 85 ) are not confined to the largest ovules studied .
It is therefore quite possible that here , as in Gnetum , the embryo deyelops in the fallen seed before germination .
The only seeds that have been available for examination which contained large and fully developed embryos were lying on the ground and had probably been in that condition for six months .
All are agreed that , of the three living genera , Ephedra is the most nearly allied to the lower Gymnosperms .
If the group is monophyletic , it is not a little remarkable that its highest members should , in the deferred development of the embryo , have reverted to conditions so characteristic of primitive plants .
Beyond the fact that in Gnetum the oospore elongates to tubular suspensor , we have little information as to the early changes following fertilisation It is probable as in Ephedra and Welwitschia , a primary suspensor with am initial cell at its tip is formed , as a result of the first division of the oospore-nucleus , though BOWER states that no terminal cell is cut off before This tubular suspensor is described as undergoing branching , a condition which has not been lnet with in Welwitschia .
After the appearance of a few anticlinal walls in the terminal cell of the pro-embryo , " " the peripheral cells of the group thus formed grow laterally along the surface of the suspensor and , dividing further by anticlinal walls , form short KARSTEN , ls93 .
, ls79 .
LAND , 1907 .
S BOWFR , 1882 ; LOTSY , 1899 .
KARSTEN , 1892 , fig. 25 ; LOTSY , 1899 .
BOWER , 1882 .
though oppeared tehave aewhose f\ldquo ; \ldquo ; witschiaemb y than these.\ldquo ; * The roup referred to this citation has , at the apex , afew cells investigation , will bring to light other points of resemblance in the structure of the embryos of Gnetum and Welwitschia .
The pro-embryo of Welwitschia has attained a high degree of specialisation as is seen in the apparently stereotyped structure of its secondary suspensor and in the marked regularity of the cell-divisions which occur in it .
In these respects it may be compared with that of the higher Coniferae , though traces of a parallel development are lacking .
Indeed , the structure of the pro-embryo provides a strong argument in favour of placing Welwitschia and the Coniferae on widely divergent lines of descent .
If the embryonic plate is correctly identified , the pro-embryo of Welwitschia bears a striking resemblance to that of Araucaria .
S Whether this is limited to the cap and the embryonic plate or extends also to the many-celled suspensor is not clear , for the origin and structure of the latter organ are not known in sufficient detail in Araucaria .
If this resemblance is not merely apparent , it is of great interest that a type of pro-embryo so different from that of other Gymnosperms should be common to genera so widely separated as Welwitschia and Araucaria .
It must be regarded as an inheritance from an exceedingly remote common ancestry , and is no doubt correctly included among the many primitive characters which are retained by the modern representatives of the group Araucarieae K. ndosperm ( Plate 30 , figs. 88 and 89 ) .
The endospel.m of a pollinated ovule in which fertilisation has not yet occurred is circular in transverse section and narrowed towards the antipodal end .
In unfertilised ovules little further growth takes place .
A new phase is soon after fertilisatiou .
This secondary growth proceeds more rapidly after the penetration of the endosperm by the first pro-embryo .
The terete form of the primary endosperm is soon lost and there results an organism elliptic in section , except at the antipodal end , where it remains terete ; the major axis of the ellipse is tangential to the surface of the rachis of the cone ; the endosperm thus acquires a persistent bilateral form which corresponds with that of other floral organs .
The most advanced endosl ) seen is about five times as broad as at the end of the period of primary growth .
Increase in length takes place mainly at the BOWER , 1882 .
, figs. 6-8 .
, .
280 .
S STRASBURGER , , Taf .
20 , figs. 66-70 ; SEWAnD and FORD , 1906 , fig. 28 , , H. SEWARD and FORT ) , 1906 , p. 395 .
PROF .
H. H. W. PEARSON : FURTHER OBSERVATIONS ON WELWITSCmA .
371 antipodal end , which for a time behaves as an organised growing point ; the cells of a massive apical group are equally meristematic ; its activity has greatly diminished by the time that the stage of fig. is reached .
* The maximum length found is three times as great as at beginning of the secondary growth .
There are also formed during this period two shoulder-like upgrowths , one on each side of the now empty bases of the embryo-sac-tubes ( fig. 89 , 1 , K ) .
It may be that these are due indirectly to the absence of any opposing pressure in the empty cells of the fertile region which lies between them .
The tissue of the endosperm in these later stages is not uniform in structure .
If it con tains one or more embryos , whether or not its upper tissues have been destroyed figure , , an axial group of cells lying beneath the deepest embryo , and having the form of an inverted pyramid , is packed with starch ( text-figure , which at this time is not abundant elsewhere .
Outside this storage region the cells differ considerably in size .
Very small cells usually form groups or bands ularly distributed through a tissue mostly composed of much larger ones ( fig. 88 ) .
Canals formed by the growth of deep-seated embryo-sac-tubes are obliterated in the course of the secondary growth of the endosperm .
Compartments whose nuclei remain free but do not produce tubes ( cf. fig. ) are not found in an endosperm in which secondary growth is well advanced ; they presumably collapse and their contents become absorbed .
Between the stage of fig. and that of the adult seed the endosperm increases until all the nucellus , except a thin brown inyesting skin and the shrivelled remains of the nucellar cone , has disappeared .
The adult seed thus has two distinct seed-coats , , the nucellar tegmen and the outer horny winged envelope formed by two opposite leaf structures .
Owing to the intercalary growth of the nucellus above the origin of the outer envelope and the depth to which the interposed penetrated by the endosperm , the crushed ument covers little more than the shrivelled remains of the nucellar cone .
and Discussion of Results .
Both male and female cones are frequently produced in great profusion .
The inflorescence is normally a compound dichasial cyme more or less modified by arrest .
In some cases it consists of a single cone .
The branching of the escence is carried further in the male than in the female .
Subfoliar inflorescences are produced with great frequency .
Macrospores and embryo-sacs are very frequently present in the pith region of the axis of the female cone .
This implies that the ovule is cauline and confirms the view already based by most authors on other characters .
In the heterotypic division of the microspore-mother-cell the spindle is at first * An antipodal meristematic cone seems to ) in netum ( LoTSY , .
PEARSON , 1906 , p. 293 .
multipolar , later bipolar .
The post-meiotic number of chromosomes is and 26\mdash ; probably 25 .
The pollen-grains frequently germinate in the micropylar tube at a considerable distal ) above the top of the nucellus .
In these cases the pollen-tube grows downwards in the fluid medium which fills the micropyle at the time of pollination .
A single macrospore-mother-cell is organised .
A condition later than synapsis is the only stage of the meiotic divisions that has been found .
Only one macrospore is functional .
Sometimes two , rarely three , pass through the early of germination .
In these cases the lowest eventually develops at the expense of the rest .
In the developlnent of the embryo-sac a large central vacuole is not formed ; this may indicate that the growth of the cytoplasm and the multiplication of the nuclei are more than is usual among the Gymnosperms .
The nuclei divide simultnneously .
The tendency to establish polar arrangement of the nuclei which is lnanifested after the first mitosis in the macrospore is not seen in later stages .
After the last simultaneous division , the sac probably contains 1024 free nuclei which are equivalent in all visible characters .
A rapid increase in the length of the sac , occurring principally in its upper part , causes a scattered distribution of the nuclei which are closely crowded after the last simultaneous divisio1l .
As a result the nuclei of the upper fourth are more widely separated from one another than are those of the rest of the sac .
The scattered nuclei of the upper fourth are sexually functional ; the more crowded nuclei of the lower -fourths , potentially equal to the rest , give rise to a sterile tissue , the endosperm .
The whole sac now becomes incompletely septate into compartments which are very in form and size .
Those of the upper fourth contain few nuclei , usually not more six ; the great majority of those of the lower three-fourths endose many nuclei\mdash ; commonly 12 or more ; a few of the latter fi.equently resemble those of the upper fourth in that they contain only six nuclei or less .
The walls of compartments of the micropylar fourth grow upwards into the nucellus as tubes into which the nuclei and most of the cytoplasm pass .
These are the embryo-sac-tubes ; the nuclei within them are functional gametes .
The sexual nuclei lie in contact with one another in the young tubes .
Rarely they fuse within the tube and form a fusion-nucleus quite similar to an original endosperm-nucleus .
They are apparently quite free in the cytoplasm of the embryo- sac-tube ; in some cases a thin layer of cytoplasm immediately surrounding the nucleus stains more deeply the rest .
Fertilisation normally occurs in the lower half of the nucellar cone not far from the line in the text-figure .
Sometimes it is effected in the upper half of the nucellar cone .
It probably rarely takes place below the level of the top of the embryo-sac .
It never occurs within the embryo-sac .
PROF .
H. H. W. PEARSON : FURTHER OBSERVATIONS .
ON WELWITSCHIA .
373 In the lower three-fourths of the sac the free nuclei enclosed in a compartment rarely undergo division .
When this does occur they divide simultaneously or , apparently , the division may be confined to a single nucleus .
All the nuclei in a compartment which contains more than six fuse and it thus becomes a uninucleate cell .
Cells formed in this way constitute the primary endosperm .
The single nucleus of each is formed by the fusion of potential gametes .
conclusion is based on the following evidence:\mdash ; 1 .
At the of fig. 21 ( the last simultaneous division of the free nuclei being completed ) , the nuclei are of the same generation\mdash ; the eleventh from the macrospore-nucleus .
2 .
They are all alike in all visible characters .
3 .
Shortly before septation commences they show differences in size and capacity ; but they all pass the same sel.ies of changes and ultimately are again all alike .
The observed differences are due to the fact that these changes commen ce at the micl.opylar end sooner than in the lower part of the sac .
4 .
When septation is completed some compartments contain few nuclei ( six or less ) ; others a larger number .
The latter are normally confined to the lower three-fourths of the sac ; the former may occur anywhere , they usually fill ffie upper fourth of the sac and are either absent from or much less abundantly developed in the lower three-fourths .
5 .
In compartments containing less than six , wherever they may be situated , the nuclei do not fuse , but remain free ; these are functional gametes .
Fusion occurs only in those nents in which the nuclei are in number ; a sterile tissue is thus formed .
6 .
In an abnormal case in which the upward growth of the sac was by a second one aboye it , compartments which normally would have contained gametes and produced embryo-sac-tubes became uninucleate endosperm cells by nuclear fusion .
7 .
Under certain circumstances no endosperm is formed at all , but , with the exception of a few which succumb to ths greater activity of their neighbours , all the nuclei remain free and every compartment produces an ] embryo-sactube of the normal type .
8 .
In two cases ( fig. 30 , A and B ) about six gametes in an advanced embryo-sactube have fused to form a nucleus in all respects like that of a primary endosperm-cell .
9 .
A small tissue found in an embryo-sac-tube and consisting of seyen cells is interpreted as an endosperm formed by such a nucleus as those referred to in S 8 .
{ 374 PROF .
H. H. W. PEARSON : FURTHER OBSERVATIONS ON WELWITSCHIA .
The cells containing single fusion-nuclei , filling the whole or almost the whole of the lower three-fourths of the sac , form a nutritive tissue or organism\mdash ; the endosperm \mdash ; which undergoes considerable increase in size during two stages of growth , the one occurring before and the other after its penetration by the first pro-embryo .
This endosperm , resulting from the fusion of gametes , differs fundamentally from the prothallus of the lower Gymnosperms .
It is here called a trophophyte .
On the germination of the pollen-grain the generative cell increases in size , becomes elongated , and moves down the tube behind the tube-nucleus .
The nucleus of the generative cell divides once , giving two male nuclei .
The tube-nucleus usually shows signs of breaking down by the time that it reaches the level of the line in the text-figure ; it eventually disappears .
In rare cases the pollen-tube passes down the nucellus to the chalazal region of the ovule .
The bi-nucleate generative cell frequently remains undivided ; it may however become broken into two uninucleate portions .
The male nuclei may either remain spherical in form or become elongated .
Both may ersist until the time of fertilisation or one becomes disorganised .
When cytoplasmic communication is established between the pollen-tube and an embryo-sac-tube , the leading female nucleus enters the generative cell within which fertilisation occurs .
A resting fusion-nucleus is formed as a result of fertilisation ; the whole or part only of the cytoplasm of the generative cell , with perhaps a small contribution from the embryo-sac-tube , forms the cytoplasm of the oospore .
Several oospores are commonly formed in the same nucellus .
The oospore elongates , especially in the direction of the embryo-sac .
rfhe nucleus moves towards the lower end of the pro-embryonal tube thus formed , and divides .
In rare cases the pro-embryonal tube grows towards the micropyle .
The division of the primary pro-embryonal nucleus is followed by the formation of a wall cutting off a small cell at its lower end .
This wall apparently arises as a centripetal rowth from the wall of the tube .
The upper cell undergoes no further division .
It increases greatly in length , at first as the primary suspensor and later as the axial cell of the secondary suspensor .
From lower cell is produced a group of initial cells , from which are derived successively : inner cortical cells ; ( b ) 16 outer cortical cells of the secondary suspensor ; ( c ) a presumed " " embryonic plate\ldquo ; of 8 cells ; ( d ) a ring of 16 cells around and slightly below the latter ; ( e ) a terminal ' ' cap\ldquo ; of 8 cells .
By the time that the series ( b ) is formed , the tip of the pro-embryo has usually reached the endosperm within which its further development proceeds .
It is possible that the later stages of the development of the embryo are passed through after the seed has become detached .
Secondary growth in the endosperm , which commences after fertilisation , is to some extent localised in the antipodal end .
Finally , all the nucellus is displaced save a thin brown tegmen and the shrivelled remains of the cone .
The adult endosperm has a bilateral symmetry .
PROF .
H. H. W. PEARSON : FURTHER OBSERVATIONS ON WELWITSCHIA .
375 The trophophyte of Welwitschia is ically so different from the prothalIus of Ephedra that these genera must be placed in two distinct groups , whose affinities are indicated by characters of subordinate importance .
What little is known of the development of the endosperm of netum leads to the conclusion that it is a trophophyte gous with that of Welwitschia .
Since this conclusion accords with other of a fairly close degree of affinity between these genera , they are placed together in the following scheme of subdivision of the group Gnetales .
Group I. Ephedroideae .
ndosperm a true prothallus bearing archegonia ; microspore on germination producing one or two prothallial cells , a ) nucleus , a wall-(stalk- ) nucleus and two gametes ; oospore-nucleus by successive divisions giving rise to a small number of free nuclei ( frequently eight ) , each of which forms a pro-embryo ; complete intraseminal development of the embryo attained before the seed falls .
Group .
Gnetoideae.\mdash ; Female gametophyte reduced to a modium containing free sexual nuclei which are very numerous ; endosperm formed as a result of gamete-fusion ; microspore , on germination , producing an evanescent prothallial nucleus , a tube-nucleus and two gametes ; oospore giving rise directly to a single embryo ; intraseminal embryo-development , or only the later stages of it , probably deferred until the seed has fallen .
Gnetum.\mdash ; The embryo-sac or only the fertile end of it unseptate until after fertilisation ; primary endosperm formed after or before fertilisation ; fertilisation occurring within the embryo-sac which is penetrated by the pollen-tube .
Welwitschia.\mdash ; Whole sac incompletely ) tate and } ) rinnaryendosl ) formed fertilisation ; felnale ganletescarl.ied by the ) tubes into the nucellal cone , where fertilisation takes The absence of the vegetative part of the ) hyte as well of the sterile tissue of the archegonium , both of which are present in Ephedra , does not hnit of explanation .
It is not improbable that a clue to the solution of this problem may yet be found in some species of Ephedra or of Gnetum whose life-histories are at present unknown .
It is indeed possible that the fusion belieyed to 0CCU among the nuclei of the jacket-cells of distach .
represents one of a series of changes which has led to the conditions now obtaining in Welwitschia .
A detailed study of every available form of both genera is certainly called for .
In the absence of any direct evidence as to the manner in which the free gametes of the Gnetoideae- have been deriyed from the intrasporic prothallus and archegonia\mdash ; and this must be assumed to have occurred\mdash ; two alternative suggestions are tentatiyely dvanced .
* In G. , etc. ; in some other ) ocies , uccording to STEN , tho cleus b } somewhat as in Ephedra .
There is a well-marked tendency towards the elimination of wall-tissue in the archegonium of the higher groups .
This is carried so far that in the Cycads and some Conifers the wall is reduced to two neck-cells .
If the reduction proceeded but little further along the same lines , an archegonium , represented merely by an oosphere and a ventral callal-cell , would result .
Snch archegonia occur not infrequently in Picea The ventral canal-cell is almost certainly to be regarded as an arrested gamete .
If the complete reduction of wall-tissue were accompanied by the reversion of the ventral canal-cell to its presumed primitive condition of a functional gamete , the archegonium of the lower gymnosperm would be reduced to two free gametes .
Although in most of the living archegoniate Spermaphytes , so far as they have been investigated , the limited number and definite Iocalisation of the archegonia indicate a high of specialisation , cases are known in which neither of these conditions is realised .
Perhaps the most remarkable of these is found Microcycas which the archegonia are very numerous and the irregularity of their distribution so marked that they are not even confined to the surface of the prothallus .
This is regarded by CALDWELL as a primitive character .
A similar tendency to increase the number of archegonia appears in Araucaria , S AgathisS\mdash ; in which as many as 60 may be organised in a prothallus\mdash ; and in Sequoia These cases also the authors are inclined to consider as primitive .
But , on the other hand , so marked an irregularity in the distribution of the archegonia as is seen in these forms is not characteristic of the lower Archegoniatae , where , however , their usually limited numbers and more or less definite localisation are no doubt largely determined by physiological conditions connected with fertilisation and , more especially , with the nutrition of the embryo .
On the evidence at present available , it is at least a possibility that , as has suggested , this tendency to increase the number of archegonia , instead of being a relic of primitive conditions , represents an advance towards Gnetoid conditions .
If the number of archegonial initials were increased to the utmost extent , there might result a sac in which each of the free nuclei present after the last simultaneous mitosis which precedes septation would represent an archegonial initial .
If further the formation of archegonial wall-tissue were completely suppressed and the ventral canal-nucleus reyerted to its primitive condition of sexuality , all the nuclei in the adult sac would be free gametes .
It is therefore of some interest to find that after the completion of the last , 1903 , p. 364 , figs. 60 and 61 .
BLACKMAN , 1898 ; COULTER and CHAMBERLAIN , 1901 , pp. , 87 , fig. 65 ; LYON , F. , 1904 ; DAVIS , 1903 , 1904 ; FERTY , 1904 .
CALDWELL , 1907 .
S SEWARD and FORD , 1906 , pp. 366 , 367 .
LAWSON , 1904 .
LAND , 1907 .
PROF .
H. H. W. PEARSON : FURTHER OBSERyATIONS ON WELWITSCHIA .
377 division of free nuclei in Welwitschia the number present is greater than that recorded for some other Gymnosperms .
In Ginkgo , $ Taxus , and E. trifurca , S the total number of nuclei present in the sac in the stage immediately preceding septation ' is stated to be about 256 .
If there has been no departure from the simultaneous character of the nuclear divisions , all these 256 nuclei are of the same generation\mdash ; the ninth in a direct line from the nucleus of the macrospore .
There are , however , several cases known in which the number at a corresponding stage is much larger , e.g. , in Pinus " " about 2000 in sempervirens over 500 , and in " " about 1000 .
' ; In Welwitschia , assuming that the mitoses are simultaneous throughout , the actual number is 1024 .
All these nuclei which must now be regarded as potentially sexual belong to the eleventh generation from the macrospore llucleus .
It is stated that in Ephedra the first division of a primary alveolus which gives rise to an results in the formation of the central cell and primary neck cells , but since this observation possibly refers to E. only , it cannot at present be interpreted to mean that the nuclei of the Welwitschia-sac are of the same generation as the oosphere of E. or of E. .
But it may be suggested that the large number of nuclei produced in the Welwitschia-sac is perhaps connected with the elimination of the prothallus and archegonium as such and their substitution by free gametes .
An alternative suggestion may be based upon the behaviour of vegetative nuclei in the prothalli of certain so-called " " apogametic\ldquo ; ferns .
Here the normal sexual act is replaced by the fusion of two nuclei of adjacent cells of the prothallus ; the cells whose nuclei behave thus may be situated anywhere in the younger parts of the plant .
Whatever may be the precise difference between the normally vegetative and the sexual nuclei of the prothallus , it may thus disappear to the extent of enabling the fusion of two belonging to the former category to produce a normal embryo .
It is possible that the development of the characters of the female gamete in all the nuclei of the prothallus , together with the suppression of wall-formation , has been concerned in bringing about the change from the conditions of the lower seed-bearing plants to those of the Gnetoideae .
The early stages of development of the Welwitschia-sac correspond very closely with those described for the lower Gymnosperms , and the changes which involved HIRASE , 1895 .
JAEGER , 1899 , p. 252 .
JACCARD , 1894 , p. 16 .
S LAND , 1904 , p. 10 .
FERGUSON , 1904 , p. 84 .
LAWSON , 1904 , p. 1 , 1907 , .
129 .
BERRIDGE , 1907 , p. 281 .
, 1907 , .
175 .
VOL. CC.\mdash ; B. 3 in the acquisition of sexual characters by all the sac-nuclei do not manifestly occur earlier than the end of the free-nuclear condition .
In Gnetum all the nuclei of this last stage are apparently already potentially sexual .
In Welwitschia they undergo certain changes in form and staining capacity in later development and it may be that here sexuality is not acquired until after septation .
But in both , the gametes are free nuclei irregularly distributed in a non-septate or incompletely septate plasmodium .
It is therefore probable that in the transition , by a gradual process or per saltum , from the sac of the lower Gymnosperm to that of the Gnetoideae , the nuclei which became endowed with sexuality were those which filled the still unseptate sac at the end of simultaneous nuclear division .
Hence the gametes in the primitive sac of the Gnetoid.type were free nuclei .
The view that the Angiosperm-sac is a derivative from the same type is advanced below .
One of the of its higher degree of specialisation is the fact that its gametes , functional and reduced , are no longer free nuclei but cells .
It remains to be considered whether the Gnetoid endosperm is unique or whether , on the other hand , it may not be regarded as representing a primitive condition , from which that of the higher seed-bearing plants has been evolved .
The former is the product of the division of many fusion-nuclei , the latter of one ; in both the fusion- nucleus results from the union of nuclei of the same generation as the functional gametes .
Angiosperms are described in which the endosperm is apparently formed without antecedent nuclear fusion ; these seem to be exceptional and rare .
The sac of Peperomia exhibits certain peculiarities , on account of which a comparison between it and that of Gnetum has already been suggested .
* In the four-nucleate stage of P. the sac shows no polarity , but the nuclei are arranged peripherally , " " like the spores of a tetrad\ldquo ; \mdash ; in fact , precisely as in Welwitschia in the corresponding stage ( fig. 16 ) .
When the mitoses are completed there are sixteen nuclei in the sac belonging to the fifth generation from the macrospore .
These are " " rather evenly distributed through the sac\ldquo ; ; one of them increases in size and becomes the nucleus of the oosphere ; of the rest , eight fuse to form the primary endosperm-nucleus .
Each successive division of this nucleus and its descendants is followed by the formation of cell-walls , as in Welwitschia .
In P. hispidula , of the sixteen sac-nuclei all but two fuse to form the definitive nucleus .
S The general resemblance of this process to that by which the Welwitschia endosperm is formed is very striking ; in both cases the constituents of the primary endosperm nuclei are of the same generation as the functional gametes ; in Welwitschia , not only is this the case , but the constituents of the fusion-nuclei are themselves potential gametes .
There is no evidence that the fusing nuclei of * COULTER and CHAMBERLAIN , 1903 , p. 90 .
The same rangement is recorded for the Penaeaceae EPHENS , .
JOHNSON , 1900 ; CANPBELL , 1901 .
S JOHNSON , 1907 , SARGANT , 1908 .
PROF .
H. ARSON : FURTHER OBSERVATIONS ON LWITSCHIA .
3 Peperomia are potential gametes , but they are probably to be regarded as the arrested representatives of nuclei which in earlier forms were potentially sexual .
CA MPBELL* regards the sac of Peperomia as among Angiosperms .
JOHNSON , on the other hand , inclines to the view that its abnormalities rather indicate a higher degree of specialisation .
As CAMPBELL has pointed out , the evidence for this view is inadequate .
Even if the plant is reduced in its vegetative parts and in its floral structure , and if , as and maintain , the crowding of the flowers into a spike is not a primitive feature , it does not follow that a departure from the " " normal ' structure in the embryo-sac is not primitive .
doubled number of nuclei cannot be attributed to reduction , and neither this nor the absence of polarity is in agreement with the suggestion of increased specialisation .
In Cypripedium , S in which there is good reason to believe that a high of specialisation has been attained , the number of sac-nuclei is just half that found in normal cases .
While irregular divisions among the sac-nuclei conceivably result from disturbing conditions in the environment , it is difficult to suppose that any ffictor other than heredity is concerned in the regular production of twice the normal number of sac-nuclei and the ol.ganisation of the oosphere in a generation once removed from that of the oosphere in the normal Angiosperm .
CAMPBELL'S view further receives strong support from the facts now established Welwitschia .
It is a justifiable conclusion , from the evidence available , that the sixteen-nucleate sac of Peperomia represents a stage intermediate between an type , with a massive primary endosperm , and the more recent eight-nucleate form , which is found in the majority of the higher Angiosperms .
The fact that the number of constituents of the fusion-nucleus varies in different species of Peperomia is favourable to this view , for in Welwitschia , a form less specialised , in which there are many fusion-nuclei in the same sac , the number of their constituents shows a wide range of differences .
In this colmection , NAWASCHIN'S observations on ; lans are of great interest .
There is no organised egg-apparatus , but the sac contains several free nuclei , any one of which may be fertilised .
KARSTLN has extended this observation to other species of Juglans .
Except for the small number of nuclei , this seems not far removed from the condition of the Welwitschia-sac .
In Aglaonem " " multiple nuclear fusion\ldquo ; frequently occurs , and it is likely that it gives rise to the endosperm-nucleus .
In the higher Angiosperms , in which specialisation of the gametophyte has proceeded *CAMPBELL , 1901 .
JOHNSON , 1900 . . .
, 1903 , pp. 90 , 91 .
, 1903 , .
674 .
further , and the oospbere appears in the fourth or third generation , the fusion-nucleus contains only two constituents .
These are normally the two polar nuclei .
But this is not constantly the case , for in ipedium , in which the sac contains only four nuclei , the constituents of tho imary endosperm-nucleus , in addition to a male nucleus , are ( , one from a synergid and the other from the antipodal region .
Of the six nuclei of the normal sac which do not participate in the fusion , one is functionaly sexual ; two others , those of the synergids , are in some cases fertilised and produce normal embryos .
Embryos , said to be apogametic , are described as originating from the antipodal cells , from the upper polar nucleus , S and from the cells of the endosl ) the last in elongata , in which the primary endosperm is not formed a fusion-nucleus .
An " " antipodal egg\ldquo ; esulting from the fertilisation of antipodal nucleus a spel m is believed to occur in some species of aster .
Clearly , then , the -nuclei sometimes are potentially sexual , and although no cases in which the polar or antipodal nuclei become normally fertilised are certainly known , it is probable that as regards sexuality they are merely more reduced than the syllergid-nuclei .
The group of three or more antipodal cells in the angiosperm-sac has acquired an absorptive function in l1lany families , but in a great number of others they appear to very little if any part in the activities of the sac , but become disorganised and eventually replaced by endosperm .
It is interesting to compare this behaviour with that of the " " othallium cells\ldquo ; of the fertile part of the sac of Gnetum nuclei of these cells are related to the functional galnetes in the same nuclei of deep-seated in the Welwitschia ndosI)fig .
; and since they pres mably belong to the same generation as the sexual nuclei , they are comparable with the antipodal and synergid-nuclei of the Angiosperm and with the " " accessory\ldquo ; nuclei of Peperomia , though ically they } ) ) ably diffel in that they normally etain a potential sexuality .
And their fate is the same as that of the non-persistent antipodal cells , the synergids and ccessory nuclei\mdash ; after fertilisation " " the fertile part of the embryo-sac . . .
is obliterated , owing to the growth of the sterile part ; its retarded rotl ) cells play no hatever .
In both cases the endosperm derived from fusion-nuclei destroys and replaces all the contents of the embryo-sac , save the embryos themselves .
1907 .
, ( \fnof ; mCotTI , tLAIN , 1903 , p. 217 ; OVERTON , 1891 , ex MANN , 1892 , 1 ) .
373 , , 1901 , it1ld CHAMBERLAIN , 1903 , , 216 , , fig. 103 .
; , 1878 , and CHANBEItLAIN , loc. , p. 221 .
S , lS99 , CHAMBERLAIN , 1903 , p. 218 .
\ldquo ; , , lJ04 .
, 1899 , , p. 110 .
PROF .
H. H. W. PEARSON : OBShRVATIONS ON LWITSCHIA .
381 These considerations suggest that the embryo-sac of the higher Angiosperm is a highly specialised structure , derived through such forms as are represented by Peperomia , Juglans , and Penaea from ancestors in which the conditions of the Gnetoid sac were more or less nearly realised .
If so , the endosperm of the Angiosperm was originally a trophophyte homologous with that of Welwitschia .
In the course of the evolution of the existing Angiosperms it has become reduced , passu , with the appearance of the functional gamete in successively earlier generations , and the consequent reduction of the number of nuclei available for fusion .
Its constitution has further been fundamentallypodified by the inclusion of a male gamete in the primary fusion-nucleus from which it is derived .
There seem to be no cases described in which more than one primary } ) nucleus is formed in the Angiosperm , while in the Gnetoideae the number is large and the fusion is preceded by septation .
This difference is no doubt closely connected with the great reduction in the number of sac-nuclei .
In Corydalis a septation followed by fusion apparently very similar to that found in the Gnetoideae is described .
Whether the fusing nuclei in this case are the descendants of a normal definitive ( fused ) nucleus or of the two free polar nuclei does not seem to be certainly established .
* In the Gnetoideae the division of the fusion-nucleus is immediately followed by wall-formation .
In the Angiosperm , on the other hand , the formation of a number of free nuclei usually precedes septation .
In many cases , however , the first division of the definitive nucleus is followed by the appearance of a wall dividing the sac into two cells , and even when this does not occur a transitory cell plate is frequently formed .
Therefore it is probable that in the ancestors of the Angiosperms the successive divisions of definitive nucleus and its descendants were followed by the formation of cell-walls as still occurs in Peperomia and in the Gnetoideae .
One of the most marked features of the sac of the Angiosperm is its pronounced polarity .
The difference between this condition and that of Welwitschia is perhaps not so great as would appear at first sight .
In the latter , as also in Gnetum , there is an inherent tendency towards a polar arrangement of its nuclei after the division of the macrospore-nucleus ( fig. 15 ) .
But if corresponding stages of development up to the time of fertilisation are compared , the sac of Welwitschia is probably not much larger than that of the average Angiosperm , while the number of nuclei rapidly becomes reater .
It is therefore possible that the approximately uniform distribution of the nuclei in the former is connected with , perhaps a direct consequence of , the fact that after the first few nuclear divisions the sac is not large enough to allow of any other arrangement .
It has been seen that nuclei in any part of the Welwitschia-sac may remain unfused and so retain their sexuality .
If this occurs more frequently in one part of the normally sterile region than in another it is , perhaps , at the antipodal end .
Since , STRASBURGER , 1880 .
STRASBURGER , 1879 , Taf .
14 , fig. 60 ; KARSThN , 1892 , p. 210 , fig. 5 ; 1893 , Taf .
9 , fig. 44 .
however , this is not of general , it is probably not connected with the ten dency to establish polarity seen in the earlier stage , but is rather due to the fict that growth of the sac is llot uniform .
The exceptional occurrence of a sac which the nuclei retain their sexuality and no endosperm is formed rather indicates that Welwitscbia is descended from forms in which all the nuclei were otential gametes at the time of fertilisation , and the conditions were such that any one of them , wherevel , might be fertilised .
Theoretically these conditions realised in the species of Gnetum studied by lack of ) Juglans , and other Augiosperms* in which etained other characters reasonably egarded as primitive , justifies the belief that the more typical angiosperm-sac has been derived from a form in which the polar arrangement did not persist beyond the bi-nucleate stage .
A considerable reduction in the number of nuclei formed\mdash ; owing to the organisation of the functional gametes in an earlier generation than in Welwitschia or Gnetum\mdash ; especially if accompanied by an in the size the sac , would be likely to result in a .
of the nuclei ather than their distribution in the cytoplasm .
If this grouping commenced mediately after the first division of the macrospore-nucleus the result would be a bipolar arrangement such as is seen in the normal Angiosperm .
If it were defel.red until the appearance of the grand-daughter nuclei , there might be four peripheral groups such as are described for the Penaeaceae .
A group of three organised cells , ally 1 ) otential gametes , would provide the materials for the evolution of the highly specialised is formed normally at the micropylar ) and , in some forms , is still present at the antipodal end .
has suggested that each of the four roups of the Penaeaceae is a ratus .
If this is it will constitute important evidence in favour of the group of the normal sac as reduced eggatus .
But on the hypothesis now advanced this does not appear to be necessary in cases .
All that is assumed is that the eight nuclei are either all potentially sexual or , in most cases , seven or less are the reduced representatives of nuclei which ossessed t character .
it is ) obable t in very many groups no such slJecialisation o in the antipodal cells as is istic of the functional eggapparatus . .
persistence in the higher Angiosperms is no doubt largely to be accounted for by the fact that they have acquired a new function and play a more or less important in the itio of the sac .
In the remarkably reduced sac of Cypripediuln an is not Whethel the organisation of antipodal cells in the Augiosperms has thy connection with a chalazogamous method of fertilisation is doubtful .
The fact that Welwitschia the generative cell may be duwll into the tissue of tlJe enode below the origin of the , 1903 , 1 .
90 , IPBELL , 1903 , , 670 . .
PROF .
H. H. W. PEARSON : FURTHER OBSERVATIONS ON LWITSCHIA .
38.3 integument , and therefore below the antipodal end of sac , does not necessarily indicate the former occurrence of a specialised lnethod of tilisatio akin to the chalazogamy of some iosperms ; it may equally be relic from a , the sac at the time of .
filled equivalent gametes , the pollen-tubes entered it at any points in its wall .
A further possibility is that the solitary case has been seen is purely accidental and due to the unusual circumstance that the pollen-tube has evaded the embryo-sac-tubes in ] nucellar cone and therefore continued its downward course .
The phenolllenon of triple fusion , known to occur in ) enera and 22 nlilies of has made more complex the problem of the of endosperm in this roup .
Diverse views as to the nature of the fusion and its phylogenetic significance have beenl advanced .
None of these is together in accol.d with the hypothesis that the endosperln of the ) itive Ariosperm was essentially homologous with that now described for Welwitschia .
No that can be considered as phylogenetically connected with the triple fusion occurs in Welwitschia .
There is at present no evidence to show that a nlale gamete plays any part in the formation of the primary -nucleus in Peperomia or in any other in which more than two of the sac-nuclei cerned therein .
It is suggested that the male nucleus was not included in the fusion which gave rise to the definitive nucleus in the itive A , and that the origin of the ) henomenon is not to be sought in the higher Gylnnosperlns .
It is rather probable that triple fusion was called into existence by the continued reduction of the number of sac-nuclei and consequently of the number available as constituents of primary endosperln- nucleus .
It may be assumed that the movement of the polar-nuclei towards one another or of the lower one towards the upper is brought bout by a mutual attraction , no doubt of a chemiotactic nature .
In some species of ] this action is sufficient to draw togethel not only the polar-lluclei ) , with , tbose of the antipodal cells and synergidae .
The second male nucleus , after its projection into the -sac , is free to lnove , and , according to observations 011 Monotropa , S is carried down to the polar-nuclei the ovenlents in an axial strand of cytoplasnl .
It is probable that its movelnent is affected by the same directive force which is potent in the case of the -nuclei .
some cases ) is a third possibility , that the male nucleus may ) ossess some of independent lnotion which enables it to make its ownl way to fusing perllaps also under the influence of a directive force .
But whetlJer the nucleus etail l the lnotility as well as the , 1908 , .
SARGANT , 1900 ; LANI ) , 1907 ; THOMAS , 1907 ; BrRRIT 1907 .
URBECK , 1902 , ClIAhlBFRLAIN , 1903 , p. 93 .
S STRASBURGER , 1900 , ex SAIt , 1900 . .
' it .
form of the antherozoid or , having lost both , is carried by the streaming cytoplasm , with or without the influence of an internuclear attractive force , the conditions prevailing in the angiosperm-sac are highly favourable to the occu-rrence of the triple fusion .
It is suggestive that when , as sometimes happens , more than one of the sac-nuclei retain their sexuality and therefore presumably exert a more powerful attractive influence upon the second male gamete , the triple fusion is prevented and a true double fertilisation occurs ; this occurs , e.g. , in , where one male nucleus fertilises the oosphere and the second may fuse with a synergid , in which case two normal embryos are formed .
* It is therefore worthy of consideration whether the triple fusion is not primarily a result of the peculiar conditions of the higher types of angiospelull-sac and confined to sacs in which they are realised .
Three suggestions have been made as to the importance of the triple fusion .
The stimulus to renewed growth is increased , and the introduction of a set of characters from the individual which stands as the male parent of the embryo may aise the nntritive efficiency of the resulting endosperm .
Further it is supposed that the increase in the number of chromosomes caused by the entrance of the male nucleus may serve to secure the degeneracy of the resulting tissue .
S In Welwitschia the primary fusion-nuclei are numerous .
If all but one were eliminated this would closely resemble in origin the definitive nucleus of Peperomia , and ceteris paribus would produce a very small mass of endosperm tissue .
Miss SARGANT points out that a similar result is achieved by the definitive nucleus of Peperomia In this case there is no evidence that the triple fusion occurs , though the difficulties of observation are great and it appears to be still possible that it may have been overlooked .
If , however , the origin of the definitive- nucleus is correctly described , Peperomia affords a most interesting comparison with Welwitschia on the one hand and with the higher Angiosperm on the other\mdash ; it suggests a raison for the triple fusion .
With the reduction in the number of the sac-nuclei , and consequently of the number of fusionnuclei , the trophophyte as established in Welwitschia is no longer an efficient endosperln .
Its efficiency is increased by the incorporation in the fusion-nucleus of a male gamete .
It is conceivable that the introduction of this new factor has not only saved from extinction a race which discarded the prothallus in favour of the trophophyte , but has also been of fundamental importance in determining the success of the Angiosperms in the present epoch .
The change from the intrasporic female prothallus retaining the sexual apparatus of the Pteridophyte to the trophophyte with free gametes , judged by its results , has been hardly less importance in the evolution of the higher plants than was the *GUIGrNARD , 1901 , and CHAMBERLAIN , 1903 , p. 216 , fig. 103 .
SARGANT , 1900 , p. 705 .
THOMAS , 1907 : SSARGANT , 1900 .
SARGANT , 1908 , p. 129 .
PROF .
H. H. W. PEARSON : FURTHER OBSERVATIONS ON WELWITSCHIA .
385 earlier advance from the conditions of the holophytic or saprophytic prothallus to those of the seed .
The latter has already commenced in the living Pteridophytes ; there is at present no evidence as to the manner in which the former was initiated .
The conclusion that the intrasporic prothallus underwent a series of reductions in the evolution of the higher archegoniate Gymnosperms is notjustified by such evidence as is at hand .
But it is possible that its replacement by the trophophyte was connected with the development of a necessity for a more active and efficient means of nourishing the embryo than it provided .
Such a necessity might through the shortening of the period comprised between pollin ation and the maturation of the seed .
It is probable that the seed appeared independently in different groups of heterosporous Pteridophytes , and it is equally likely that , under stress of similar conditions , more than one group of Gymnosperms discarded the prothallus for a nutritiye organism originating in gamete-fusion\mdash ; a possibility which must be considered in discussing the relationships between the Gnetoideae and the Angiosperms .
After the establish- meant of the trophophyte it seems that reduction and specialisation set in .
The oosphere was organised in successively earlier generations , and the number of nuclei available for fusion became successively smaller .
From a condition in which many fusion-nuclei were formed ( as in Welwitschia ) there was derived a sac\mdash ; such perhaps as is now seen in Peperomia\mdash ; in which only one was produced .
The endosperm resulting is apparently no longer an efficient organism .
Still further reductions in the direction of the conditions of the higher Angiosperms may be supposed to have involved a further lowering of the efficiency of the endosperm until the triple fusion appeared .
The direct descent of the Angiosperms from the Gnetales has been mnch discussed , and although LoTbY and other authorities have pronounced against it , it still finds a place among current theories .
If the essential homology of the endosperm in the former group with that of the Gnetoideae be admitted , there is established between them a closer degree of affinity than has hitherto been suspected .
The absence of archegonia and the freedom of the female gametes within the sac , together the total disappearance of the tative portion of the gametophyte , are no less significant of the existence of a line of separatio1l between the lower Gymnosperms ( including Ephedra ) on the one hand and the Gnetoideae and the Angiosperms on the other .
The reduction of the male and sexual cells in Gnetum and Welwitschia to conditions almost identical with those prevailing in some Angiosperms and the small size of the male cells in Welwitschia add further emphasis to this distinction .
Some small differences in the development of the microsporangium and in the dehiscence of the are not likely to be of primary importance .
The relations of the floral structures in the two groups have already been discussed , and these seem to place considerable difficulty in the way of regarding any Angiosperm as directly descended from Welwitschia and perhaps equally from Gnetum .
The living Gnetales form a group of undoubted antiquity .
* PEARSON , 1906 .
If the endosperm of the Angiosperm is morphologically of the same nature as that of Gnetum and Welwitschia , it must follow that , in a broad sense , the Angiosperms and the Gnetoideae are derived from the same stock .
* It is , however , improbable that the two groups are on the same line of descent .
Gnetum and Welwitschia are rathel the last representatives of a race which separated from the main Angiosperm line after the replacement of the intrasporic prothallus by the trophophyte .
An attempt has been made to show that the floral structures of Welwitschia may conceivably be derived from a type of strobilus similar to that which was characteristic of Bennettites .
Very little is known about the ovule of this extinct type and , indeed , it is not certain that unfertilised ovules have yet been found .
But certain characters of ovules which do not contain the normal dicotyledonous embryo are suggestive of a closer affinity with Welwitschia than any that can be deduced from a possible resemblance of external floral characters .
The evidence is however at present quite inadequate and the following comparison is purely tentative : Plate 30 a section is represented passing through two ovules in which the cavity enclosed by the " " thin-walled nucellus\ldquo ; is nearly or entirely filled by " " large , rounded , and thin-walled cells This tissue is " " ularly , though characteristically , traversed by bands , which under very favourable conditions are seen to be made up of small \ldquo ; The author does not regard this tissue as a prothallus , because there is no indication of the presence of archegonia .
He inclines rather to the opinion that it is a massive undifferentiated lnbryo such as that of Ginkgo .
Against this it may be urged that in its comparatively undifferentiated conditionS the embryo of Ginkgo , while it fills the cavity of the archegonium , has a transverse diameter much smaller than that of the prothallus ; in a transverse section it would not " " nearly , if not entirely , fill the seed cavity And further , a section of the embryo of the living Ginkgo is not traversed by irregular bands of small-celled tissue There is , of course , also the possibility that the sections , being transverse , have passed through the protha ] below the level of the archegonia .
An alternative suggestion is that these ovules are , as WIELAND at first gests , unfertilised or contain only undeveloped pro-embryos and that the tissue in question is not a prothallus but a trophophyte of the same character as that which occurs in Welwitschia .
This would account for the absence of archegonia and also for the presence of the irregular bands of small-celled tissue which is a constant character of the Welwitschia endosperm ( fig. 88 ) .
If this is the true nature of the sterile tissue of the Bennettitean sac , the possibility of a enetic connection between Bennettites and Welwitschia , which has already been discussed , will be very greatly increased .
*Cf .
and CHAMBERLAIN , 1903 , pp. 286 , 287 .
, 1906 , pp. 124 , 125 . .
cit. S COULTER and CHANBEltLAlN , 1901 , fig. 35 ; LYON , H. L. , 1904 , Plate 33 , fig. 8 ; Plates 30 , 32 .
Loc .
PROF .
H. H. W. PEARSON : FURTHER OBSERVATIONS ON WELWITSCHIA .
387 Perhaps the most astonishing feature of the life-history of Welwitschia is the acquisition by the apparatus connected with the female gametes of a behaviour which is so constantly characteristic of the male .
The cells containing the female gametes not only have departed from the state of quiescence which marks the female organ in all but the lower Algae and Fungi , but , in the entrance of the female nucleus into the male cell and in the very large part played by its cytoplasm in the construction of the oospore , we have characters which are probably unparalleled in the vegetable kingdom .
There is an important difference between the formation of haustorial tubes by the micropylar end of the sac in Gnetum* and in some Angiosperms and the growth of the embryo-sac-tubes in Welwitschia .
In the latter the tubes are formed by the membranes of the compartments containing the fertile nuclei , while in the former the nucellus is penetrated by the wall of the embryo-sac itself In no species of Gnetum yet described are there any structures of like character .
The apex of the archegonium of grows upwards into the neck as the time for fertilisation approaches and may pass through it into the pollen-chamber .
S A growth of a similar kind is recorded for Fortunei While there is a certain similarity between this and the growth of the embryo-sactubes of Welwitschia , it can hardly be supposed that they are phylogenetically connected .
At the same time the high degree of organisation to which the latter have attained makes it improbable that they have originated de novo in Welwitschia .
Since they invariably attain nearly to that region of the nucellar cap in which fertilisation normally occurs before the ovule is pollinated , it seems probable that their importance in the economy of the plant is primarily associated with a shortening of the period between pollination and fertilisation\mdash ; in other words , that they are of the nature of an adaptation to extreme xerophytic conditions .
was probably the view entertained by STRASBURGER when he wrote of them , " " docb scheint mir dies eine specielle Anpassung zu sein , auf die kein zu grosses Gewicht zu legen \ldquo ; When these words were written it was not known that the formation of the tubes was preceded by a partial septation of the fertile end of the sac .
A similar septation of the lower part of the sac normally occurs in Gnetum , usually after fertilisation .
In G. , however , the process is advanced a stage and precedes fertilisation .
In Welwitschia it has been extended from one end of the sac to the other .
Even if the tubes have arisen in the manner suggested it is impossible that the antecedent septation can have originated merely as an adaptation to external conditions .
* KARSTEN , 1893 ; , 1906 , p. 288 .
HOOKER , 1863 , p. 39 .
The curious tubular cells in the upper part of the sac of Gnetum ( LOTSY , 1903 , , figs. 2 , 3 , 4 , ' Taf .
10 , fig. 1 ) , which Dr. oTSy believes to be parthenogenetic embryos , arc very suggestive of inoipient embryo.sac-tubes .
S BERRIDGE and SANDAY , 1907 , p. 132 .
COKER , 1907 , fig. 7 .
STRASBURGER , lS72 , p. 295 .
The separation of Welwitschia from the immediate neighbourhood of Ephedra and the establishment of affinities , real if not close , with Gnetum , might be anticipated from the geographical distribution of the three genera .
Ephedra is not represented in Africa , south of the Mediterranean region .
Its absence from Equatorial and Southern Africa is striking in view of its wide east and west distribution in the warm temperate ions of the Northern Hemisphere and its development on both sides of the Equator in America .
Gnetum , with an equally extensive east and west range in the Tropics , is represented in Africa by two climbing species confined to the Guinea region , viz. , .
Bucholzianum in the Cameroons*and G. Africanum on the Quetta Mountains , about 120 miles from the coast at St. Paul de Loanda .
The littoral strip of desert , of whose flora Welwitschia forms the most striking feature , is a continuation northwards of the arid belt which fringes the western coast of extratropical South Africa .
On north and north-east it is bounded by the southern extension of the Guinea region and on the east by the woodland formationS of the mountainous districts of inner Damaraland .
It has been shownS that , in spite of the remarkable development of endemic and peculiar forms in the desert strip , there are , nevertheless , indications that its flora is allied to that of the Acacia region , which itself is probably connected with that of Guinea through Ovamboland and the lower and intermediate slopes of the Angola Plateau .
The geographical evidence is therefore not opposed to the view that the floras to which the African Gneta and Welwitschia respectively belong are descended from the same phytogeographical stock .
At the same time , the physical raphy of the area of together with the unique characters of the dominant elements of its flora , indicate that the conditions which have led to its differentiation have persisted through an enormously long period .
The fact that Welwitschia diverges so widely in its external characters , as probably also in the details of its reproduction , from species which , for geographical reasons , may be supposed to be more nearly related to it than any other living plants , is therefore not remarkable .
If forms intermediate between Gnetum Gnemon and Welwitschia still exist , they are to be looked for in G. and G. , or in other species occurring in the same phytogeographical region , and at present undiscovered .
[ It is due to the author to state that , during the last few months , he has been conducting a Sladen Botanical expedition in different parts of Africa , and has therefore had no opportunity of noticing the most recent literature bearing on the subject of this paper.\mdash ; A. C. S. ] Botanical South Afi.ican College , Cape Town .
* ENGLER , 1908 .
BOLUS , 1905 .
RENDLE , 1899 .
S , 1907 ( b ) .
PEARSON , 1907 ( b ) , 1908 .
PROF .
H. H. W. PEARSON : FURTHER OBSERVATIONS ON SCHIA .
389 NDIX I. My colleague , Prof. A. BROWN , has kindly calculated the number of nuclei in the sac , after the completion of the last of the series of simultaneous mitoses , from the median longitudinal section shown in fig. 30 .
The method employed , which is of general application to such cases , is here given .
A median line ( see fig. SO ) is drawn , and divided parts ; ordinates are drawn at each of the points of section and measured ; call The area of the section is taken as and the volume of the solid as these being elementary approximations .
If be the number of nuclei in the section , is the mean surface to one nucleus .
If the distribution is uniform we may assume to be the mean volume to nucleus , hence the number of nuclei in the volume is , i.e. , ' The measurements giye cm .
, and the 's giving sq .
cm .
, Also number in the volume is ( 185 ) 0 .
0.4515 3 .
0.9404 0.9404 Giving a number , 1015 .
APPENDIX Il .
METHODS .
As far as possible the fixing was done in the field at the time of collection .
The winged perianth was in all cases removed from the more advanced ovules before they were placed in the solutions .
The apex of the cone-axis with its attached bracts and very young flowers was fixed as a whole .
In cases in which the fixation had to be deferred , sometimes for a period as long as three hours , the results compare more favourably than might be expected with those yielded by the material fixed when collected .
For the stages of the embryo-sac preceding fertilisation , by far the best results were obtained with a solution of picric acid and mercuric chloride in a 5-per-cent .
solution ( alcoholic ) of glacial acetic acid .
* The use of a solution of chromic and osmic acids in 4-per-cent .
glacial acetic acid ( aqueous ) was hardly less successful , especially for embryo-sac-tubes , pollen- tubes , and sexual nuclei .
The bulk of the material very satisfactorily fixed in a solution of chromic and acetic acids , according to a formula used by Mr. I. B. POLE EVANS , .
:\mdash ; Chromic acid ( l-per-cent .
solution in distilled water ) 200 Glacial acetic acid . . . .
1 Distilled water . . . . .
50 Several other reagents were also employed , , saturated solution of mercuric chloride in l-per-cent .
acetic acid ; acetic alcohol ( 30 : 70 ) ; Carnoy 's solution ; mixtures of absolute alcohol and formaline and aqueous solutions of the latter .
Except for the purposes of comparison the material preserved in these has not been studied .
None of them appear to yield results as good as those obtained with the first three formulae .
Great difficulty was experienced in washing material after fixation .
Apparatus was prepared for immersing it in the river at Haikamchab , but as this was quite dry it could only be washed by frequent changes of water or alcohol .
It was then taken up to 70 per cent. alcohol or Calberla 's fluid .
Later , in the laboratory , it was again washed in changes of alcohol .
The preparations were made by the paraffin-method .
Most of the drawings are made from sections 8 thick .
The haematoxylins ( Heidenhain , Ehrlich , and Delafield ) alone or in combination with alcoholic eosin or erythrosin , combinations of diamant-fuchsin and licht-griin and of cyanin and erythrosin were the principal *CHAMBERLAIN , 1906 , p. 20 .
CHAMBERLAIN , 1906 , p. 327 .
PROF .
H. H. W. PEARSON : FURTHER OBSERVATIONS ON LWITSCHIA .
391 stains used .
Diamant-fuchsin and been especially valuable in the investigation of the pollen-tubes and embryo-sac-tubes .
The former is the only stain that has made clear the structure of the starch grains .
Difficulties , which have.not yet been overcome , have attended the investigation of the nuclear changes intervening between the stages of the microspore and the pollen-grain .
The wall of the microspore is already so thick that stains penetrate very slowly , and this , no doubt , also applies to the reagents .
* The hardness of the wall and the small size of the grain has made it impossible to obtain good sections .
The only method which has yet been attended with any degree of success has been to stain in bulk for seyeral days in Ehrlich 's haematoxylin , and afterwards to extract the excess of colouring matter by means of acid alcohol .
But the nuclear structures are always obscured by the opacity of the wall , even after clearing .
The drawings have been made under the ; in all cases Zeiss achromatic objectives and Huygenian eyepieces have been used .
BIBLIOGRAPHY .
ARBER , E. A. N. , and PARKIN , J. , 1907 .
" " On the Origin of the Angiosperms ' Journ. Linn. Soc vol. 38 , p. 29 .
BALFOUR , J. B. , 1901 .
Report of the British Association ( Glasgow meeting ) , p. 819 , London .
BENTHAM , G. , and HOOKER , J. D. , 1880 .
'Genera Plantarum , ' vol. 3 , London .
BERRIDGE , E. M. , 1907 .
" " The Origin of Triple Fusion ' New Phytologist , ' vol. 6 , p. 279 .
BERRIDGE , E. M. , and SANDAY , E. , 1907 .
" " Oogenesis and nbryogeny in Ephedra \ldquo ; 'New Phytologist , ' vol. 6 , p. 127 .
BERTRAND , C. E. , 1878 .
" " tude sir les teguments seminaux des Vegetaux phanerogames ' Ann. des Sci. Not .
, ( 6 ) , vol. 7 , p. 63 .
BLACKMAN , V. H. , 1898 .
" " The Features .
of Fertilisation and related Phenomena in Pinus ) estr.is , \ldquo ; 'Phil .
Trans B , vol. 190 , p. 395 .
BOLUS , H. , 1905 .
( Sketch of the Floral Regions of South Africa ' Science in South Africa , ' p. 199 , Cape BOWER , F. O. , 1881 .
" " On the Germination and Histologry of the Seedlings of mirabilis 'Quart .
Journ. Micr .
Sci vol. 21 , p. 15 .
Idem , 1881(b ) .
" " On the further Development of Wehuitschia \ldquo ; loc. cit. , p. 571 .
Idern , 1882 .
" " The Germination and Embryology of Gnetum Gnemon loc. cit. , vol. p. 278 .
* PEARSON , 1906 , figs. 12-16 .
I am indebted to Dr. B. DAYDON JACKSON , See .
L.S. , for tt photograph of this paper .
CALDWELL , O. W. , 1907 .
" " Microcycas Calocoma 'Botanical Gazette , ' vol. 44 , p. 118 .
CAMPBELL , D. H. , 1900 .
" " Studies on the Araceae ' Annals of Botany , ' , p. 1 .
Idem , 1901 .
" " The Embryo-sac of Peperomia .
cit. , vol. 15 , p. 103 .
ldcrn , 1903 .
Studies on the Araceae .
cit. , vol. 17 , p. 665 .
CHAMBERLAIN , C. J. , 1905 .
( Methods in Plant Histology , ' ed. 2 , Chicago .
, 1906 .
" " The Ovule and Female Gametophyte of Dioon 'Botanical Gazette , ' vol. 42 , p. 327 .
COKER , W. C. , 1907 .
" " Fertilisation and Embryogeny in Cephalotaxus Fortunei ' Botanical Gazette , ' vol. , p. 1 .
COULTER , J. M. , and CHAMBERLAIN , C. J. , 1901 .
" " Morphology of Spermatophytes part 1 , New York .
Idenv , 1903 .
'Morphology of Angiosperms , ' New York and London .
DAVIS , B. M. , 1903 .
" " The Origin of the Archegonium 'Annals of Botany , ' vol. 17 , p. 477 .
ldem , 1904 .
" " The Relationships of Sexual Organs in Plants ' Botanical Gazette , ' vol. 38 , p. 241 .
Idem , 1904-1905 .
" " Studies on the Plant Cell The American Naturalist , ' vol. p. 367 .
EICHLER , A. W. , 1889 .
" " Gnetaceen Die Not .
Pfl .
-Fam .
, ' Tei12 , Abt .
1 , p. 116 .
ENGLEI A. , 1907 .
'Syllabus der Pflan zenfamilien , ' ed. 5 , Berlin .
lden ?
, 1908 .
" " Gnetaceae Afiicanae , \ldquo ; 'ngler 's Bot. Jahrb vol. , p. 5 ] 9 .
ARMOUR , J. B. , and DIGBY , L. , 1907 .
" " tudies in Apospory and Apogamy in Ferns 'Annals of Botany , ' vol. , p. 11 .
FERGUSON , M. , 1904 .
" " ibutions to the Knowledge of the Life-history of Pinus , etc. ' Washington Acad. Sci. Proc vol. , A. W. , 1906 .
" " The Morphology and dling Structure of the Geophilous cies of Peperomia , together with some Views on the Origin of the Monocotyledons 'Annals of Botany , ' vol. , p. 395 .
HILL , T. G. , 1906 .
" " On the Seedling Structure of certain Piperales .
cit. , p. 113 .
HIRASE ' , S. , 1895 .
" " Etudes sir la fecondation et l'embryogenie du biloba ' Coll. Sci. Tokyo , p. 103 .
HOLFERTY , G. M. , .
" " The Archegonium of 'Botanical Gazette , ' vol. , p. 106 .
HOOKER , J. D. , 1863 .
" " On Welwitschia , a new Genus of Gnetaceae 'Linn .
Soc. Lond. Trans vol. , p. 1 .
JACCARD , P. , 1894 .
'Recherches Embryologiques sir l'Ephedra ' Lausanne .
, L. , 1899 .
" " Beitrage sir Kenntniss der Endospermbildung und sir Embryologie yon Taxus ' Flora , ' vol. , p. 241 .
JOHNSON , D. S. , ' ' On the Endosperm and Embryo of pellucida ' Botanical Gazette , ' vol. , p. 1 .
Miss E. L. STEPHENS , B.A. , has kindly furnished me with an abstract of this paper .
PROF .
H. H. W. PEARSON : FURTHER OBSERVATIONS ON WELWITSCHIA .
393 KARSTEN , G. , 1892 .
" " Beitrag sir Entwickelungs-Geschichte einiger Gnetum-Arten , \ldquo ; ' Bot. Zeit vol. , p. 205 .
Idem , 1893 .
" " Sir Entwickelungs-Geschichte der Gattung Gnetum 'Cohn 's Beitr . .
Biol. .
Pfl vol. , p. 337 .
LAND , W. J. G. , 1904 .
" " Spermatogenesis and Oogenesis in ' Botanical Gazette , ' vol. , p. 1 .
Idem , 1907 .
" " Fertilisation and Embryogeny in Ephedra \ldquo ; loc. cit. , vol. LAWSON , A. A. , 1904 .
" " The Gametophytes , Archegomia , Fertilisation , and Embryo of Seqnoia ) \ldquo ; ' Annals of Botany , ' vol. , p. 1 .
Idern , 1904 .
" " The Gametophytes , Fertilisation , and Embryo of Cy yptor .
cit. , p. 417 .
, 1907 .
" " The Gametophytes and Embryo of the Cupressineae , with special reference to \ldquo ; loc. cit. , vol. 21 , 281 .
LIGNIER , O. , 1903 .
" " La Fleur des Gnetacees est-elle intermediaire entre celle des Gymnospermes et celle des Angiospermes ?
\ldquo ; ' Bull .
Soc. Linn. Norm ( 5 ) , vol. 7 , p. 55 .
, J. , 1848 .
'An Introduction to Botany , ' ed. 4 , London .
LOTSY , J. P. , 1899 .
" " Grosser Morphology of Production of Gnetum Gnemon , ' Ann. Jard .
Bot. But ( 2 ) , vol. , p. 46 .
Idern , 1903 .
" " Parthenogenesis bei Gneturn , BRONGN ' Flora , ' vol. 92 , p. 397 .
LYON , F. , 1904 .
" " The Evolution of the Sex Organs of Plants 'Botanical Gazette , ' vol. , p. 280 .
, H. L. , 1904 .
" " The Embryogeny of Ginkgo 'Minnesota Botanical Studies , ' vol. 23 , p. 275 .
McNAB , W. R. , .
" " On the Development of the Flowers of mirabilis , HOOK .
fil .
' Linn. Soc. Lond. Trans vol. , p. 507 .
MANN , G. , 1892 .
" " The Embryo-sac of Myosurus minimus , L. : a Cell Study 'Bot , Soc. Edinb .
Trans vol. , p. 351 .
MIyAKE , K. , 1903 .
" " On the Development of the Sexual Organs and Fertilisation in Picea excelsa 'Annals of Botany , ' vol. , p. 351 .
MOTTIER , D. NI .
, 1904 .
'Fecundation in Plants , ' Carnegie Inst. , Idem , 1907 .
" " The Development of the Heterotypic Chromosomes in Pollen Mothera cells ' Annals of Botany , ' vol. , p. 309 .
OPPERMAN , M. , 1904 .
" " A Contribution to the Life-history of Aster 'Botanical Gazette , ' vol. , p. PACE , L. , 1907 .
" " Fertilisation in Cypripedium ' Botanical Gazette , ' vol. , p. 353 , PARLATORE , P. , 1869 .
" " Gnetaceae in DE CANDOLLE , ' Prodromus , ' vol. , fasc , 2 , p. S47 .
PEARSON , H. H. W. , 1906 .
" " Some Observations on .
HOOK .
fil 'Phil .
Trans , vol. 198 , p. 265 .
VOL. CC.\mdash ; B. 3 .
H. W. ARSON : FURTHER OBSERyATIONS ON WELWITSCHIA .
PEARSON , H. H. W. , 1907 .
" " The Living Welwitschia ' Nature , ' vol. 75 , p. 536 . .
Some Notes on a Journey from Walfisch Bay to Windhuk Kew ' ( 1907 ) , p. 347 .
Idenr , 1908 .
" " Botanical Excursion in the Welwitschia Desert\ldquo ; ( Abstract ) , Report of British Association ( Leicester Meeting ) , p. 685 , London . .
B. , lS99 .
'Catalogue of the African Plants collected by Dr. FRIEDRICH WELWITSCH in 1853-61 , ' vol. 2 , part 1 , London .
RICHARD , A. , 1825 .
'Nouveaux de Botanique et de Physiologie tale , ed. 3 , Paris .
, 1900 .
" " Recent Work on the Results of Fertilisation in Angiosperms Annals of Botany , ' vol. , p. 689 .
Idern , 1908 , " " The uction of a Race of Primitive giosperms loc. cit. , vol. , p. 121 .
SCOTT , D. H. , 1900 .
'Studies in Fossil Botany , ' London .
WARD , A. C. , and , S. O. , 1906 .
" " The Araucarieae , Recent and Extinct Phil. , vol. 198 , p. 305 .
SOKOLOWA , C. , 1891 .
'Naissance de l'endosperme dans le sac embryonnaire de Gymnospermes , ' Moscou .
, H. CRAF .
" " Bennettites gibsonianus 'Annals of Botany , ' vol. , p. PHENS , L. , 1908 .
" " A preliminary Note on the Embryo-sac of certain Penaeaceae loc. cit. , vol. 22 , p. 329 .
BURGER , 1872 .
'Die Coniferen und die GnetaceeIl , ' Jena .
Idem , 1879 .
'Die Angiospermen und die Gymnospermen , Jena .
Idem , 1880 .
'Zellbildung und Zelltheilung ed. 3 , Jena .
THOMAS , E. N. , 1907 .
" " Some Aspects of Double Fertilisation in Plants 'Science Progress in the Twentieth Century , ' vol. , p. 420 .
VAN , P. , ' ' Anatomy comparee de la Fleur femelle et du Fruit des Cycadees , des Coniferes et des Gnetacees 'Ann . .
Sc. Not .
Bot. ' ( 5 ) , vol. 10 .
WARMING , E. , 1895 .
'A Handbook of Systematic Botany ' ( Eng. Trams London .
, G. R. , 1906 .
Amel.ican Fossil Cycads , ' Carnegie Institute , Washington .
Idem , 1908 .
" " Accelerated Cone-growth in Pinus ' Am .
Journ. Sci vol. 25 , p. 102 .
DELL , W. C. , 1901 .
" " The Vascular Structure of the 'Flowers ' of the Gnetaceae , 'Annals of Botany , ' vol. , p. 766 .
ldem , ] ) .
" " The Structure and the Morphology of the Ovule\mdash ; an Historical Sketch loc. cit. , vol. 18 , p. 57 .
* hliss E. L. SNEPHENS , B.A. , has kindly furnished me with an abstract of this paper .
PROF .
H. H. W. PEARSON : TIONS ON WELWITSCHIA .
395 DESCRIPTION OF PLATES .
KEY .
Figs. 1- 7 Plate 22 8-18 , , 23 , , 45 , , 27 29-86 . . . .
, , , , , , 29 Figs. 81-89 . . . . .
Plate 30 .
A. -floral -sacs .
Fig. 1.\mdash ; Outline of median longitudinal section ] ) portion of a female cone showing the positions of the in and , and the summit of that shown .
) ' strand .
Fig. 1 , , c.\mdash ; The embryo-sacs indicated in ; .
sterile tissue of cone-axis ; -sac .
B. -cell } Fig. 2 , ; Fig. 7 , ; the Fig. 2.\mdash ; Part of a section a ginm -nuclei are in the } ) } ) has of the rotyl ) divisioit .
) iderlnis ; .
outer and innel tapetum ; cell .
Fig. 3.\mdash ; Prophase of heterotypic division .
Fig. 4.\mdash ; Telophase of heterotypic Fig. 5.\mdash ; Equatorial-plate-stage of homotyl ) diyisions .
Fig. 6.\mdash ; Anaphase .
Fig. -plate formation .
* C. Structures in ollinat i. S-ll .
) Figs. 9 , 10 , ) ; -sac-tube ; outer layer of collt } ; inner layer of integument ; .
integument ; .
nucellus ; ) . .
pollen-tube .
Fig. .\mdash ; Photograph of male cones in situ showing ) rojecting smens ( PEARSON , 1907 ) .
Fig. .\mdash ; Photograph of female cones in sitn ' micropyles , some of which are tipped with drops of the fluid .
* For later stages of microspore and pollen-grcin , 1906 , figs. .
ctio young ated ) ovule passing through Part , of ovule passing rough iument of nucellar cone .
ction t lower part ylar tube and top of showing ) rains a pollen-tubes .
1 ) .
-ccll , -sac .
Figs. 12-23 . ?
1 ? .
-wall ; lLncellus .
-mother-cell with nucleus in ) ) of heterotypic division ( 1 ) .
( Cf .
Pearson , 1906 , in ction .
-nucleus in ( two successive sections ) , showing about ( tendency towards a ) olar arrangement of , nudei .
-sac with nudoi as a old ) great development of Cltoles a the nuclei ) in mitosis .
( .
L9 .
{ ( older lnst , in whid } all the nuclei in . .
30 .
full llnmber of free nuclei .
ylar end of the sac he planes of ) tiot these ) ) ents will form i11to lclei will ) ' section of the same sac .
The ) of ) similar to those in fig. 22 . .
smaller nuclei .
If none of these six will all ftl.ltt uninucleate cells of the PROF .
H. H. } .
PEARSON : ON LWITSCHIA .
397 E. The of the after the ( Figs. 24-31 .
) Fig. 27 , ; the rest , endosperm ; embryo-sac-tube ; macrospore-wall ; nucellus ; .
starch grains .
Fig. 24.\mdash ; Part of longitudinal section showing last stages of wall-formation and free sexual nuclei .
PEABSON , Fig. 25.\mdash ; A later stage in which the outgrowth of embryo-sac-tubes has commenced .
One compartment contains five nuclei .
PEARSON , Fig. 26.\mdash ; A median longitudinal section through the tissue of the nucellar cone lying immediately above the embryo-sac and into which young embryo-sac-tubes have penetrated .
One tube contains five free sexual nuclei which lie close together near the tip .
Fig. somewhat tangential longitudinal section through the upper part of the embryo-sac and the superincumbent nucellar tissue showing embryo-sactubes , one of which is growing downwards .
Fig. 28.\mdash ; An older -sac-tube containing two sexual nuclei ' some distance behind the tip .
Fig. 29 .
\mdash ; A similar tube four nuclei , a little contracted .
Fig. 30 , A and B.\mdash ; Very exceptional cases in which the nuclei in embryo-sac-tubes have fused .
The fusion-nuclei closely resemble that of an endosperm-cell .
( Cf .
fig. 35 .
) Fig. 31.\mdash ; A more advanced tube in which the nuclei are separated .
F. Thoe Lower Three-fourths of the Sac after Septation the Primary Endosperm .
Figs. Figs. and 39 , ; the rest , containing a fusion-nucleus ; other lettering as in previous section .
Fig. 32.\mdash ; Same sac as fig. 24 .
Multinucleate compartments distinctly marked out , but cell-walls not yet visible .
PEARSON , Fig. 33.\mdash ; Same sac as fig. 25 .
Cell walls .
In some } ) nents the nuclei have already fused .
A single nucleus in mitosis .
( Cf .
PEARSON , 1906 , fig. ) Fig. one multinucleate compartment the eight nuclei are in simultaneous division ( five seen in the section ) ; one contains a fusion-nucleus ; the two small cells have been formed by the division of a fusion-nucleus .
Fig. 35 , .\mdash ; Stages in the fusion of the nuclei to form uninucleate endosperm-cells ; in , a primary endosperm-nucleus is in n1itosis .
Fig. 36.\mdash ; A longitudinal section through a part of the primary endosperm in which nuclear fusion is completed .
In one cell is seen a stage in the formation of a cell-plate .
Fig. 37 , A and , an embryo-sac-tube containing a seven-celled tissue ( end .
) identified as an endosperm formed from a fusion-nucleus such as those shown in fig. 30 .
, a group of small cells from the normal endosperm of the same ovule drawn to the same scale .
Fig. 38 , A and , a portion of a longitudinal section through the antipodal end of the same sac as fig. 26 , showing a compartment with free nuclei surrounded by uninucleate endosperm-cells .
, A deep-seated compartment in which the two nuclei have not fused ( only one appears in the section ) ; its position is shown at in fig. Fig. 39.\mdash ; A longitudinal section passing through the junction between two embryo sacs in the same ovule situated one above the other .
No nuclear fusion has yet occurred in the upper one ; in the lower , all the compartments contain fusion-nuclei except a few situated at the side where the pressure is least ( see p. 354 ) .
0 .
of the Pollen-grain and Growth of the Pollen-tube .
Figs. Fig. 41 , ; fig. 47 , ; the rest , .
endosperm ; .
exine ; generative cell ; .
chalazaltissue formed below the integument by intercalary growth ; .
intine ; .
integument ; .
nucellus ; perianth\ldquo ; ; tube ; -nucleus .
Fig. in early stage of germination .
Fig. 41.\mdash ; Part of pollen-tube showing e]ongated generative cell .
The tube bends sharply at " " \ldquo ; to enter the summit of the nucellus .
Fig. 42.\mdash ; A cell of the nucellar cone showing compound starch grains .
Fig. 43 , A and B.\mdash ; Pollen-tube shortly after entering the nucellus , showing generative cell and tube-nucleus .
Figs. 44 , 45.\mdash ; Two successive sections through a pollen-tube showing the nucleus of the generative cell in mitosis .
The tube-nucleus is seen in fig. 45 .
Fig. 46.\mdash ; A pollen-tube after the mitosis of the nucleus of the generative cell .
Fig. 47.\mdash ; Outline of longitudinal section of an ovule showing the course of a pollen- tube into the chalazal region ( f ) .
Fig. 48.\mdash ; The generative cell whose position is indicated in fig. 47 .
Fig. 49.\mdash ; A generative cell .
A deeply stained mass seen in the next section , lying just in front of the generative cell , represents the tube-nucleus .
The leading male nucleus is smaller than the one behind it ( as is usualy the case ) and here stains more deeply .
PROF .
H. H. W. PEARSON : FURTHER OBSERVATIONS ON WELWITSCHIA .
399 Fig. 50 , A and B.\mdash ; Two successive sections of a pollen-tube showing two approximately equal male nuclei and a tube-nucleus in process of absorption .
Fig. 51.\mdash ; A generative cell .
The second nucleus is greatly elongated .
Fig. generative cell .
The leading nucleus is much smaller than the hinder one , stains very deeply , shows little internal structure and has an irregular outline .
It is regarded as being in process of disintegration .
H. Meeting of the Embryo-sac-tube and , and Fertilisation .
Figs. 53-67 .
Figs. 53 , ; fig. 64 , ; the rest , aborting female nucleus ; other letterin as in preceding section .
Fig. 53.\mdash ; Part of a longitudinal section through the nucellus showing an embryo-sactube and a pollen-tube approaching one another .
Fig. 54.\mdash ; Parts of the tubes of fig. 62 enlarged .
The female nucleus shows an anterior papilla .
A mass of cytoplasm containing a degenerating nucleus , five sections away from this one , is probably the second half of the generative cell .
Fig. 55 .
\mdash ; Embryo-sac-tube and pollen-tube meeting ( the tip of the pollen-tube destroyed in preparation ) .
The female nucleus shows an anterior papilla , and is ensheathed in a thin layer of deeply staining cytoplasm .
The second half of the generative cell , containing a male nucleus in normal condition , is present in the succeeding section .
Fig. 56.\mdash ; A -nucleate generative cell and , in front of it , allucleus which is identified as 9 .
The lower part of the tube is continuous below with a callal in which lies a celled pro-embryo .
This is interpreted as a case in which one of th lower felnale nuclei has been fertilised before the leading one .
This is unusual .
Fig. 57.\mdash ; An embryo-sac-tube and pollen-tube in contact .
The female nucleus has entered the cytoplasm of the pollen-tube .
The male nuclei in the generative cell are in normal condition .
A second female nucleus can be identified in this embryo-sac-tube below the limit of the figure .
Fig. .58.\mdash ; Three of four successive sections of the same embryo-sac- and pollen-tubes ( one section between and is not shown ) .
In , the leading female nuoleus is within the generative cell ; the second female nucleus within the cytoplasm of the pollen-tube and a tangential section of the generative cell are seen in shows a third female nucleus still in the embryo-sac-tube , and some distance below that of B. The second male nucleus has not been identified .
Fig. 59.\mdash ; A large generative cell containing an elongated nucleus ; the second male nucleus has disappeared , unless it is represented by the deeply stained line in the middle of the generative cell .
In front is a female nucleus perhaps still within the embryo-sac-tube .
generative cell containing male and female nuclei is shown on a larger scale in fig. Fig. 61.\mdash ; Male and female nuclei almost in contact within the generative cell .
The male nucleus shows an anterior depression .
The position of the aborted male nucleus is probably shown by a small mass of deeply stained particles in the hinder part of the generative cell .
Figs. 62 , 63.\mdash ; Two successive sections through a nucellus , showing embryo-sac-tube and pollen-tube in communication .
In fig. 63 , male and female nuclei within the generative cell ; in fig. 62 , a second female nucleus in the embryo-sac-tube .
A third female nucleus , probably belonging to the same tube , is seen in another section .
Fig. generative cell containing a male and female nucleus in the act of fusing .
Fig. ( i5.\mdash ; Male and female nuclei in process of fusion .
The nucleoli are distinct .
Part of the nuclear substance stains more deeply than the rest .
The oospore-cytoplasm shows a similar differentiation ; its outline is irregular and there is yet no trace of a cell-wall .
An abortive female nucleus is partially embedded in the cytoplasIn of the Fig. 66.\mdash ; A more advanced oospore with cell-wall .
Nucleoli not yet fused .
Fig. 67.\mdash ; An oospore .
The cell-wall is not yet completely formed .
An aborting female nucleus is seen at I. The Pro-embryo .
Figs. Figs. 68-72 , fig. 75 , ; fig. 86diagrams ; the rest , .
presumed embryonic plate ; .
terminal of initial cells ; .
inner cortical ring ; .
inner cortical tier ; .
outer cortical ring ; .
primary suspensor ; other lettering as in preceding section .
Fig. 68.\mdash ; Young pro-embryo with nucleus in mitosis .
Fig. 69.\mdash ; A larger pro-embryo , with nucleus in a later stage of mitosis .
No cell-plate is visible .
Fig. 70.\mdash ; The nuclear division is completed .
The transverse wall is apparently formed centripetaUy .
Fig. 71.\mdash ; A two-celled pro-embryo .
Fig. 72.\mdash ; A two-celled pro-embryo , with nucleus of terminal cell in mitosis .
Fig. 73.\mdash ; A three-celled pro-embryo .
Fig. 74 .
\mdash ; A pro-embryo with a terminal group of four initial cells .
Fig. 75.\mdash ; An initial cell of the stage of fig. 74 in mitosis . .
H. H. W. PEARSON : FURTHER OBSERVATIONS ON LWITSCHIA .
401 Fig. .
\mdash ; A -celled -embryo with four initial cells , on the top of which lies a plate of four cells , which on division ( fig. 77 ) give rise to the inner cortical tier of eight cells Fig. 78 , A and B.\mdash ; The four , initial cells of fig. 77 lave divided longitudinally to form a group of eight , of four are seen in A and two in B. Fig. 79.\mdash ; A later stage , showing two initial cells , and above them two of a of cells resulting from the division of the initial cells in ressb in divide to give the cells of the outer cortical ring ) .
Fig. cells of the tiel have comnlellced to cells of the emmbryonic plate re ished ( ) , one of three initial cells is in division , a result of which a third embryonic cell will be formed .
The section oblique , three , of the inner cortical ring cells are seen ( also in other figures ) .
Fig. 81.\mdash ; Two cells of the inner cortical tier , considerably ated , are seen in longitudinal section ; the inner ends of two others cut transversely .
Two cells of the outer cortical ring are shown and , below them , two cells of the lower ring ( x ) which have been separated by from the terminal initial cells ; the section passes through three ) the cap-cells .
Fig. 82.\mdash ; A later stage .
The cells of the inner ring ( four shown ) are now much elongated ; those of the outer cortical ring ( two ) have comnlenced to grow .
The lbryonic plate , lower ring and cap , are each repl.esented by two cells .
Fig. 83.\mdash ; A tangential section of a stage a little more advanced tlJalltllat of , passing through the cap , the lower ring , outer cortical ring , and three cells of the inner cortical ring .
( It is possible that at least two of the inner cells of the tier lnarked " " \ldquo ; may belong to the plate , and not to the lower ) ( Cf .
fig. ) Fig. S4.\mdash ; An oblique section through a stage hardly more need than that of fig. 83 .
cells of the embryonic plate .
fig. ) and four of the cap are shown .
Fig. 85.\mdash ; A somewhat ique section of tho most need 1 ) -embryo found , surrounded by tissue .
cells of the outer cortical have elongated considerably since the stage of fig. 84 .
Fig. 86.\mdash ; Diagrams of transverse sections passing through the -embryo of 85 , at the levels , C. Fig. 87 .
Transverse section through the secondary suspensor at the of 1 ) in fig. 85 .
: The ) Figs. 88 , 89 .
) Fig. gitudinal section through an endosperm after the entrance of a mbryo .
Nuclei containing more than one nucleolus are probably primary fnsion-nuclei which have not yet divided .
Fig. 89.\mdash ; A of ontlines changes in size and form of embryo-sac and endosperm at successively older ; transverse line indicates the junction of fertile and sterile regions .
-sac-tubes not shown .
sh .
shoulders of endosperm .
i1nit of nucellus .
sac of fig. 21 .
of figs. 25 and 3.3 .
, , , , fig. 35 , , D. , , , , , , , , 36 .
a more advanced sac yet unpollinated .
of and 40 ; pollinated but yet unfertilised .
The position of the deep-seated sexual compartment of is shown at and are other sexual compRl.tnJents immersed in the endosperm .
latel after fertilisation .
A pro-embryo has not yet reached the endospel.m .
a very advanced stage containing several embryos .
cavity formed by the latter ( see text-figure , p. 365 ) .

